At the November 1967 “Conference on the Applications of Simulation Using GPSS” it seems unlikely that anyone was wondering if the conference would still be occupying a big hotel in 2017. Conferences persist for many reasons, but a technical conference like WSC has to remain relevant to users, vendors, researchers and consumers (not just hotels) to survive. If our kind of simulation vanished, then so (eventually) would WSC. What is required for simulation to “remain relevant” for the next 50 years? Without fear of having to answer for my crimes in 2067, I boldly speculate on what SHOULD matter for the next 10-20 years, if not the next 50, with a focus on our strength: dealing with uncertainty.

The conventional approach to model construction for simulation is to focus on a single model and follow a more or less structured development cycle. Why we put in twice the time and effort to develop two models rather than one? The answer lies in the fact that like most greedy heuristics, short-sightedness at the beginning may be much more costly in the end. This talk will champion the cause of the pairs-of-models (perhaps families of models) with discussion of multiresolution modeling. We show how the pair-of-models approach leads to better results overall than construction of a complex model followed by a simpler model developed subsequently by necessity under stress when complexity overwhelms. Benefits include the ability to perform mutual cross-calibration, avoiding the usual difficulties in harmonization of the underlying ontologies as well as ability to better reconcile and correlate predictions of referent system outcomes.

This paper gives a personal perspective on the evolution of discrete-event simulation—concentrating on how it moved from having an image from the 1950s through the 1980s of being a “brute force programming effort” and as a problem-solving “method of last resort” to today’s status where simulation enjoys “considerable scientific respect.” These days, simulation is often the solution “method of choice” and has gained much “scholarly respect.” This evolution changed simulation’s reliance on the use of “ad hoc methods” of solution on “early digital computers” to using simulation software systems containing “science-based methods” of solution on “modern day computers.” Remember: Simulation today has considerable scientific respect, and this respect is the result of the evolution of simulation.

The Winter Simulation Conference (WSC) is the leading international forum for disseminating recent advances in computer simulation. WSC also provides an unmatched occasion for interactions between simulation practitioners, researchers, and vendors working in all disciplines and in the academic, governmental, industrial, and military sectors. In this paper we discuss key aspects of WSC’s evolution over the past fifty years. The discussion is based on our examination of all WSC Proceedings papers published between 1968 and 2016, which collectively document much of the history of simulation and WSC. We gather and summarize interesting facts and figures about WSC authors and their Proceedings papers so as to gain insights into conference dynamics and the interconnections between notable authors and between highly cited papers. We extract relevant information from the Web of Science, Scopus, and Google Scholar databases; and we present network visualizations of the interconnections between authors and between papers.

This paper discusses the Origins and Early Years (1967-1974) of the Winter Simulation Confer-ences. Summary information is given for each of the conferences in that interval, with expanded discussion included for the 1967, 1968, and 1969 formative conferences. The last of these Early Years conferences leads into the sequential paper in this History of the Winter Simulation Confer-ence session, which deals with the Winter Simulation Conference Renaissance Period (1975-1982).

This paper discusses the history of the Winter Simulation Conference (WSC) during the period 1975–1982. This includes the collapse of the WSC in 1975, the rebirth of WSC in 1976, and the subsequent annual conferences and other significant WSC events for the period of 1976 through 1982. This was a period of great change for the WSC, with an emphasis on developing procedures to insure the long-term continuity and success of the conference.

Discrete-event simulation first emerged in the late 1950s and steadily grew in popularity to become the most frequently used of the classical Operational Research techniques across a range of industries and users. The leading advances in the evolution of discrete-event simulation software came from the United Kingdom and the USA and this presenter was engaged for some 30 years with its development and use. The presentation reviews that history as a first-hand account, specifically in the United Kingdom.

For 50 years, the Winter Simulation Conference (WSC) has brought together researchers and practitioners from the field of discrete-event simulation. This paper discusses the demographics of the authors of the many thousands of papers that have appeared in the WSC proceedings over that time span. From its origins as a “regional” conference whose participants hailed primarily from the United States, we shall see that the WSC has evolved into a truly top-flight, international conference.

This paper presents the history of the Winter Simulation Conference (WSC) for the time period of 1983 – 1992. This was a healthy time period for WSC as conference attendance was great, exhibits were added, proceedings became hardback, program tracks were added, the PhD. Colloquium initiated, and the Twenty-Fifth Anniversary of WSC was celebrated in 1992. This article discusses all of the accomplishments for this “Coming-of-Age Period”.

In this paper, we consider the history of the Winter Simulation Conference (WSC) from 1993–2007, a period characterized by growth, consolidation, and innovation. We examine developments in the WSC program including rapid proliferation of new tracks and mini-tracks to match the interests of WSC attendees. Our essay also considers the impact of technological advancements. With the launch of www.wintersim.org in 1995, the website soon became the main vehicle for dissemination of information to conference participants. Additionally, it enabled the development of the online paper-management system for submission, review, revision, and final delivery to the publisher of all papers in the Proceedings. The website also led to significant changes in how the Proceedings was published and archived. Lastly, we survey developments in the WSC administration concerning the WSC Board of Directors structure, conference financing, new conference venues (e.g., the decision to take the conference international), and novel collaborations.

In this paper we review the history of the Winter Simulation Conference (WSC) during 2008–2017, a period characterized by financial stability, continued growth, and inroads into the new age of simulation. In particular, we trace the modernization of the Conference as well as its expansion outside of the United States for the first time.

This paper offers a concise history of simulation output statistical analysis during the last six decades. Given the space limitations and historical perspective, we focus on the creation of the main concepts and methodologies that shaped the area, and proceed with a brief description of their developmental stages. We direct most of our attention to mean and quantile estimation, especially for steady-state simulations, since the bulk of the literature has been in this area, but briefly mention other topics like density estimation.

Simulation optimization has had a long and illustrious history closely tied with the 50 years of the Winter Simulation Conference (WSC). We touch upon the historical developments of the field, highlighting research progress and their interactions with WSC. Specific areas covered include ranking & selection methods, stochastic approximation and gradient estimation, response surface methodology, and sample average approximation. We discuss the interplay between research and practice, including software developments. We conclude with some reflections on the state of the field.

Statistical efficiency has been a focus of research since the inception of discrete-event simulation modeling and analysis, with origins perhaps twenty years before the first Winter Simulation Conference. We review important work in the design of simulation experiments, variance reduction through dependence structures, and efficient rare-event simulation. The focus is on the early developments, although some recent innovations also receive mention.

In stochastic simulation, input modeling refers to the process of identifying and selecting the probability distributions, called input models, from which are generated the random variates that are the source of the stochastic variation in the simulation model when it is run. This article reviews the history of the development and use of such models with the main focus on discrete-event simulation (DES).

Random number generators were invented before there were symbols for writing numbers, and long before mechanical and electronic computers. All major civilizations through the ages found the urge to make random selections, for various reasons. Today, random number generators, particularly on computers, are an important (although often hidden) ingredient in human activity. In this article, we give a historical account on the design, implementation, and testing of uniform random number generators used for simulation.

Random variate generation is a fundamental aspect of simulation modeling and analysis. The objective of random number generation is to produce observations that have the stochastic properties of a given random variable. To this end, methods and algorithms have been developed to generate random variates that are accurate (representative of the target distribution) and computationally efficient. This paper presents a history of random variate generation including distribution sampling methods used prior to the introduction of digital computers, as well as the evolution of random variate generators for continuous and discrete distributions and stochastic point processes.

The evolution of computer simulation software until the mid-1980’s is subsumed by descriptions of the history of simulation programming languages. Since that time, the entire complexion of simulation model design, development, execution, and sustainment has undergone a radical transition. The transition to a large degree stems from technology advances in hardware and software coupled with the increasing expectations of simulation modelers and end users. This study, covering the evolution in its entirety, represents an initial perspective on the transition based on an examination and analysis of the Winter Simulation Conference Archive and a partial set of simulation software surveys published in OR/MS Today. The results characterize the modeling and simulation software evolution since the mid-1980’s in terms of newcomers, endurers, fads, fades, trends, and trajectories. Prominent among the conclusions are that commercial firms are driving the major advances and the marketplace is quite volatile.

Originating in the 1970’s, the parallel discrete event simulation (PDES) field grew from a group of researchers focused on determining how to execute a discrete event simulation program on a parallel computer while still obtaining the same results as a sequential execution. Over the decades that followed the field expanded, grew, and flourishes to this day. This paper describes the origins and development of the field in the words of many who were deeply involved. Unlike other published work focusing on technical issues, the emphasis here is on historical aspects that are not recorded elsewhere, providing a unique characterization of how the field was created and developed.

This paper gives the history of verification and validation of discrete-event simulation models as seen through the eyes of its authors and their experiences. The history is divided into three time periods: the early era covering years up to 1970, the awareness era covering the years of the 1970’s and 1980’s, and the modern era covering the years of 1990 to the present.

During the past half-century simulation has advanced as a tool of choice for operational systems analysis. The advances in technology have stimulated new products and new environments without software standards or methodological commonality. Each new simulation language or product offers its own unique set of features and capabilities. Yet these simulation products are the evolution of research, development, and application. In this paper we interpret the historical development of simulation modeling. In our view simulation modeling is that part of the simulation problem-solving process that focuses on the development of the model. It is the interpretation of a real production (or service) problem in terms of a simulation language capable of performing a simulation of that real-world process. While “interpretation” is in the “eyes of the beholder” (namely us) there are some historical viewpoints and methods that influence the design of the simulation model.

This paper discusses founding of the Computer Simulation Archive at North Carolina State University Libraries, obtaining the initial contributions to the archive, establishing endowments to support the archive, and forming the archive advisory committee.

The development history of the Computer Simulation Archive is described from its inception in 1998 to the present. An overlap of visions among the creators produces an asset for the simulation community and the North Carolina State University Libraries. Collections donated over this period have produced impressive growth. Usage statistics show a steady increase in access, and the simulation endowment has nearly tripled over the past six years. The commitment of the North Carolina State University Libraries staff and the strong, consistent support of the simulation community are the key factors in this record of success.

This paper discusses the importance, historical significance, and prestige of the Computer Simulation Archive (https://d.lib.ncsu.edu/computer-simulation/) at North Carolina State University Libraries, including its importance to the field of simulation and its place in the broader context of research library archival collections.

The history of simulation and combat modeling associated with the military, and with military actions, is as old as the human race. Since humans began competing for resources, those involved have used simulation techniques of some form to better understand conflict and increase their chances of procuring favorable outcomes in those conflicts. This paper provides a historical perspective on the use of simulation in United States military planning, analyses, and training. This perspective starts by describing the simple table-top, thought exercises and proceeds through the development of ever more sophisticated methods to arrive at the modern United States military simulation environment, from laptop applications, through large-scale simulation models, to the advanced distributed simulation architectures currently in use.

In this paper we have not attempted to produce any kind of systematic review of simulation in healthcare to compete with the dozen (at least) excellent and comprehensive survey papers on this topic that already exist. We begin with a glance back at the early days of Wintersim, but then proceed, in line with the theme of this special track, to reflect on general developments in healthcare simulation over the years from our own personal perspectives. We include some memories and reflections by several pioneers in this area, both academics and healthcare practitioners, on both sides of the Atlantic. We also asked four current simulation modelers, who all specialize in healthcare applications but from very diverse perspectives, to reflect on their experiences. We endeavor to identify some common or recurring themes across the years, and end with a glimpse into the future.

Manufacturing systems incorporate many semi-independent, yet strongly interacting processes, usually exhibiting some stochastic behavior. As a consequence, overall system behavior, in the long run but also in the short run, is very difficult to predict. Not surprisingly, both practitioners and academics recognized in the 1950’s the potential value of discrete event simulation technology in supporting manufacturing system decision-making. This short history is one perspective on the development and evolution of discrete event simulation technology and applications, specifically focusing on manufacturing applications. This assessment is based on an examination of the literature, our own experiences, and interviews with leading practitioners. History is interesting, but it’s only useful if it helps us see a way forward, so we offer some opinions on the state of the research and practice of simulation in manufacturing, and the opportunities to further advance the field.

Nuclear medicine is a subspecialty of radiology that uses advanced technology and radiopharmaceuticals for the diagnosis and treatment of medical conditions. Procedures in nuclear medicine require the use of radiopharmaceuticals, are multi-step, and have to be performed under strict time windows constraints. These characteristics make the scheduling of patients and resources in nuclear medicine challenging. In this work, we integrate DEVS and CPLEX, a mathematical programming optimization software, to develop a simulation-optimization scheduling methodology for nuclear medicine clinics. We report on computational results of the new model based on a real clinic, historical data, and both patient and management performance measures. The results show that new methodology provides on average an increase of 3% on patient throughput and a decrease of 20% on patient waiting time over a scheduling policy that was used in the clinic in the past.

The transformation of innovative research ideas to production systems is highly dependent on the capability of performing realistic and reproducible network experiments. In this work, we present a network testbed consisting of container-based network emulation and physical devices to advocate high fidelity and reproducible networking experiments. The testbed integrates network emulators (Mininet), a distributed control environment (ONOS), and physical switches (Pica8). The testbed (1) offers functional fidelity through unmodified code execution in emulated networks, (2) supports large-scale network experiments using lightweight OS-level virtualization techniques and capable of running across distributed physical machines, (3) provides the topology flexibility, and (4) enhances the repeatability and reproducibility of network experiments. We validate the testbed fidelity through extensive experiments under different network conditions (e.g., varying topology and traffic pattern). We also use the testbed to reproduce key results from published network experiments, such as Hedera, a scalable and adaptive network traffic flow scheduling system.

High-performance Computing (HPC) systems have gone through many changes during the past two decades in their architectural design to satisfy the increasingly large-scale scientific computing demand. Accurate, fast, and scalable performance models and simulation tools are essential for evaluating alternative architecture design decisions for the massive-scale computing systems. This paper recounts some of the influential work in modeling and simulation for HPC systems and applications, identifies some of the major challenges, and outlines future research directions which we believe are critical to the HPC modeling and simulation community.

The following paper presents the technique that can be used to produce climatic scenarios at urban scale with a spatial resolution of 10 m based on the results of the global climate models for the different RCP climate scenarios. To make the dynamic scaling process, we use the well-known mesoscale model WRF-Chem (NOAA, USA) to produce meteorological and air quality information at different scales. We start at a first level that covers all of Europe with a spatial resolution of 25 km, until it reaches the city level with a resolution of meters, which is simulated with the model MICROSYS-CFD. To show its use, 2011 was used as reference year and 2030, 2050 and 2100 as future years, with two possible scenarios RCP 4.5 and RPC 8.5. The expected impacts on wind conditions and NO2 concentrations are shown in two European cities: Madrid and London.

The transition to a "smart city" necessitates an increase in interdependencies between critical infrastructures and information technologies. Moreover, such interdependencies are across multiple domains. However, these interdependencies expose critical infrastructures to cybersecurtiy threats. Furthermore, the availability of domain-specific simulators everywhere motivates the need for federation of interoperable cybersecurity and cyberphysical testbeds to validate cybersecurity threat resiliency. This paper presents some key issues and challenges in accomplishing such a federation of testbeds. While there are multiple modeling and simulation approaches in specific domains, none of these works address the challenges of federating across multiple domains such as federation between cyberphysical testbed and cybersecurity testbed to enable validation of cybersecurity resilience. We outline a reference architecture, DEFT (feDerate tEstbeds For cybersecuriTy) with design considerations that stem from the key issues highlighted.

Accurate and efficient calculation for expected values is challenging in finance as well as various disciplines. In general, expected values can be written as high-dimensional integrals. Monte Carlo simulation is an indispensable tool for calculating them, but it is notoriously known for its slow convergence. For spherical distributions, this paper proposes a variance reduction technique and investigates its applications in finance. By using polar transformation, the expected value is written as an integral, and the innermost integral is with respect to the radius and the outermost integral is with respect to the unit sphere. The spherical Monte Carlo estimator is the average of function values of some random points generated by lattice. We consider Value-at-Risk and expected shortfall calculation under heavy-tailed distributions and demonstrate the superiority of the proposed method via numerical studies in terms of variance, computation time, and efficiency.

In this work a multiple criteria simulation optimization method previously developed in our research group was applied to the experimental optimization of a 3D printed prototype. Both, simulation and prototyping, share the objective to provide as much information as possible about a product before its actual manufacturing. The prototype is an interlocking device that can be assembled without fastening devices or substances and that can be reassembled onto different planar assemblies. The prototype’s design considered two conflicting criteria simultaneously: maximal flexural strength and minimal mass. The method mentioned previously allowed the manipulation of a set of design variables to identify configurations with the best possible balances among both criteria. This method consists of an iterative framework based on experimental design and Pareto efficiency conditions. The aim of this work is to show the potential of the method beyond its initial intended use in simulation to approach truly experimental work.

To study external effects from supply chain on a firm's internal operations, it needs to construct a supply chain model. It may have sufficient data to build a detailed model of its operational processes. Since it has no control or visibility over the external entities, it is difficult to construct the rest of the supply chain model at the same level of detail. This paper proposes a hybrid model of the supply chain using integration design. It allows the supply chain to be captured at the broad view through system dynamics, and the target firm to be represented using agent-based model. The proposed model is therefore able to capture different behaviors of the same system, e,g, strategic planning of the whole supply chain, and detailed operational process of individual entities. To demonstrate the applicability of the hybrid model, we evaluate the model through a case study of disruption management policies.

Simulation is experimentation with a model. The behavior of the model imitates some salient aspect of the behavior of the system under study and the user experiments with the model to infer this behavior. This general framework has proven a powerful adjunct to learning, problem solving, and design. In this tutorial, we focus principally on discrete-event simulation—its underlying concepts, structure, and application.

In simulation engineering, a system model mainly consists of an information model and a process mod-el. In the fields of Information Systems and Software Engineering (IS/SE) there are widely used standards such as the Class Diagrams of the Unified Modeling Language (UML) for making information models, and the Business Process Modeling Notation (BPMN) for making process models. This tutorial presents a general approach how to use UML class diagrams and BPMN process diagrams at all three levels of model-driven simulation engineering: for making conceptual simulation models, for making platform-independent simulation design models, and for making platform-specific, executable simulation models. In our approach, object and event types are modeled as stereotyped classes and random variables are modeled as stereotyped operations constrained to comply with a specific probability distribution, while event rules/routines are modeled both as BPMN patterns and in pseudo-code.

Open Science is the practice of making scientific research accessible to all. It promotes open access to the artefacts of research, the software, data, results and the scientific articles in which they appear, so that others can validate, use and collaborate. Open Science is also being mandated by many funding bodies. The concept of Open Science is new to many Modelling & Simulation (M&S) researchers. To introduce Open Science to our field, this paper unpacks Open Science to understand some of its approaches and benefits. Good practice in the reporting of simulation studies is discussed and the Strengthening the Reporting of Empirical Simulation Studies (STRESS) standardized checklist approach is presented. A case study shows how Digital Object Identifiers, Researcher Registries, Open Access Data Repositories and Scientific Gateways can support Open Science practices for M&S research. The article concludes with a set of guidelines for adopting Open Science for M&S.

Simulation models often have many input factors, and determining which ones have a significant impact on performance measures (responses) of interest can be a difficult task. The common approach of changing one factor at a time is statistically inefficient and, more importantly, is very often just incorrect, because for many models factors interact to impact on the responses. In this tutorial we present an introduction to design of experiments specifically for simulation modeling, whose major goal is to determine the important factors often with the least amount of simulating. We discuss classical experimental designs such as full factorial, fractional factorial, and central composite followed by a presentation on Latin hypercube designs, which are designed for the complex, nonlinear responses typically associated with simulation models.

Conceptual modeling is the abstraction of a simulation model from the part of the real world it is representing; in other words, choosing what to model, and what not to model. This is the most difficult, least understood, but probably the most important activity in a simulation study. In this tutorial we explore the definition, requirements and approach to conceptual modeling. First we ask ‘where is the model?’ We go on to define the term ‘conceptual model’, to identify the artefacts of conceptual modeling, and to discuss the purpose and benefits of a conceptual model. In so doing we identify the role of conceptual modeling in the simulation project life-cycle. The discussion then focuses on the requirements of a conceptual model, the approaches for documenting a conceptual model, and frameworks for guiding the conceptual modeling activity. One specific framework is described and illustrated in more detail.

This paper aims to discuss four key elements imperative to conducting an effective simulation study and how they impact the progress of the study.

• The Clue: determining when and why to use simulation and what issues will be addressed • The Cash: understanding the financial costs and the impact of the project • The Commitment and the Courage: the importance of having a team committed to the endeavor and having the courage to make hard decisions so that the project will be successful

Each of these key factors are critical to starting a successful project and keeping it on track towards proposing effective solutions for the problems the model was designed to address.

How can you make your first project, and every project, successful? Modeling can certainly be fun, but it can also be quite challenging. You want your first and every project to be successful so you can justify continued work. Unfortunately a simulation project is much more than simply building a model -- the skills required for success go well beyond knowing a particular simulation tool. A 30 year veteran who has done hundreds of successful projects shares important insights to enable project success. He also shares some cautions and tips to help avoid common traps leading to failure and frustration.

System Dynamics (SD) can be employed for qualitative and quantitative modelling. There are important tools and methods within SD that can be easily accommodated within qualitative modeling, also known as Soft Operational Research or problem structuring method. While traditional stocks and flows are the basic components of quantitative SD modeling, quantitative SD modeling shares many commonalities, e.g. empirically driven, thorough testing, and critical focused to outputs, with traditional simulation methods and quantitative Operations Research tools. This tutorial informs novice modelers on the aspects to consider when they want to use SD as a qualitative and quantitative modeling method. In any approach employed, the use of SD modeling needs to be grounded in relevant literature from the perspective employed, qualitative or quantitative.

This paper introduces an open-source, cross-platform object called Tao to the simulation community for teaching and research. Tao is simple enough to begin using within minutes, but is extensible and interoperable with powerful statistical, graphical, animation, and simulation software, as well as internet-of-things hardware. Tao runs locally or remotely in a web browser, freeing it from OS constraints to integrate simulation engines directly into application-specific websites. Modeling examples presented here include a chemical reaction, time-bound activity sequences, and an embedded simulation for real-time load balancing. To illustrate using Tao for research, Hyden’s, single-run, global optimization algorithm was implemented. This extension enables Tao to optimize the design of dynamic systems such as queueing networks and supply chains in a single run. To demonstrate Tao’s extensiblity, the innovations of pending-intervals and a bioproduction stochastic scheduling algorithm were implemented.

The IEEE 1516-2010 - High Level Architecture (HLA) for distributed simulation is growing in a variety of application domains due to its capabilities to enable the interoperability and reusability of distributed simulation components. However, the development of simulation models based on the HLA standard remains a challenging task that requires a considerable effort in terms of both time and cost. This paper provides an introduction tutorial on developing HLA-based simulations using the HLA Development Kit (DKF) framework. The tutorial guides developers through the necessary steps for defining and creating an HLA-based simulation, and explains how the HLA elements can be easily managed by using the DKF’s services. The effectiveness of the DKF is proven by its concrete exploitation in the context of the Simulation Exploration Experience (SEE), a project led by NASA and which involves as partners several U.S. and European Institutions.

The concept of sustainable development (SDEV) is a topic of increasing significance in management decision making. SDEV is managed based on the triple-bottom line approach which stresses the importance of achieving a balance between economic, environmental and social impacts. In the context of management decision making, this implies that operational and strategic decisions in an organization must not be limited to the fulfillment of KPIs associated with productivity alone, but should also include metrics that are associated with the environment and society. Modeling & simulation (M&S) lends itself towards evaluation of the three, often competing, metrics. There are several M&S approaches like Discrete-event and System Dynamics; which of the existing techniques is the choice for modelling SDEV? Or, is a combined hybrid approach a better solution? The tutorial explores such questions related to the methodological aspects of M&S for SDEV analysis, and discusses the challenges for modeling such complex systems.

DEVS is a popular formalism for modelling complex dynamic systems using a discrete-event abstraction. At this abstraction level, a timed sequence of pertinent ``events'' input to a system (or internal, in the case of timeouts) cause instantaneous changes to the state of the system. Main advantages of DEVS are its rigorous formal definition, and its support for modular composition. This tutorial introduces the Classic DEVS formalism in a bottom-up fashion, using a simple traffic light example. The syntax and operational semantics of Atomic (i.e., non-hierarchical) models are introduced first. Coupled (i.e., hierarchical) models are introduced to structure and couple Atomic models. We continue to actual applications of DEVS, for example in performance analysis of queueing systems. All examples are presented with the tool PythonPDEVS, though this introduction is equally applicable to other DEVS tools. We conclude with further reading on DEVS theory, DEVS variants, and DEVS tools.

While the high performance computing (HPC) community is increasingly relying on the use of simulation to co-design and optimize HPC interconnects, the community lacks a coherent set of practices that should be followed when validating these simulations. Validating simulations of the HPC interconnect is a multi-phase process starting from the selection of representative communication patterns, then accurately configuring the simulated network, followed by carefully designing the set of experiments, and finally documenting the outcome for reproducibility. In this paper, we present a set of recommended practices that should be followed for each of these phases during the validation process. The end result, while complying with these guidelines, should be a validated interconnect simulation that is able to make accurate performance predictions for the simulated HPC interconnect network and that convinces the community about the correctness of the model.

From understanding our distant past to building systems of future, useful simulations demand "efficient models". Standing in the way is the twofold challenge of restraining complexity and scale of models. We describe these traits in view of component-based model development. We substantiate the roles complexity and scale play in view of modeling formalisms. We propose semi-formal modeling methods, in contrast to formal, are suitable for qualifying/quantifying model complexity and scale. For structural abstractions, we use class and component models. For behavioral abstractions, we use activity and state machines models. Furthermore, we consider these traits from the vantage point of having families of component-based models. We exemplify the concept and approach by developing families of DEVS models in the CoSMoS framework supporting DEVS-based activity and state machines models that persist in relational databases across multiple model development sessions. We conclude by discussing future research directions for real-time and heterogeneous model composability.

Communication networks are constantly evolving with new technologies providing greater quality, resiliency, and security to the data traversing current networks. In this ever-changing field, simulation provides an avenue for examining the traffic within new or existing networks. In simulating a network, characteristics and metrics of the topology may be derived without interfering with the existing framework or incurring an immediate hardware or software cost. The popular network simulator ns-3 is an effective tool for studying these network behaviors. This talk presents an overview of ns-3, discussing its design as a discrete event network simulator and its capabilities. Code snippets will be examined, demonstrating how to configure a network topology in simulation, generate packet traffic to traverse the simulated network, visualize the network behaviors, and glean metrics from the simulation. A subproject of ns-3, Direct Code Execution (DCE), is also described, demonstrating a mechanism for deploying real-world applications within the simulation.

Traffic is one of the most important aspects of modern life, both in the developed world and in the developing world. Analysis of traffic systems through modeling and simulation is an essential tool for city planners to optimize traffic flow. Microscopic discrete-event simulation is a natural and powerful method for such analysis. Characteristics of traffic systems that need to be modeled include inter-arrival times, car-following behavior, lane-changing, turning, road structure, intersection structure, and traffic light timing. There are many challenges that must be confronted as well, including the calibration and validation of traffic models, scaling models to larger traffic systems, modeling the continuous nature of traffic in a discrete-event environment, and optimization of traffic system characteristics. An emerging challenge for traffic simulation is the incorporation of autonomous vehicles into traffic systems as not only traditional vehicles, but also as cooperative agents as well.

The energy and power consumed by computing applications have long been important concerns in mobile systems and have recently become of great interest in high performance and cloud computing. To date, only a limited amount of work has considered power consumption in parallel and distributed simulation systems. A variety of options to reduce power consumption in these systems are discussed, suggestive of directions for future research in this increasingly important area.

This paper provides simulation practitioners and consumers with a grounding in how discrete-event simulation software works. Topics include discrete-event systems; entities, resources, control ele-ments and operations; simulation runs; entity states; entity lists; and their management. The imple-mentations of these generic ideas in AutoMod, SLX, ExtendSim, and Simio are described. The pa-per concludes with several examples of “why it matters” for modelers to know how their simulation software works, including discussion of AutoMod, SLX, ExtendSim, Simio, Arena, ProModel, and GPSS/H.

The representation of hybrid systems has shown to be one of the greatest challenges in Modeling & Simulation. While discrete event systems can be represented without error, continuous models rely on approximations based on numerical methods. Given the large variety of these solvers an unified representation has been elusive due the lack of an universal formalism that can describe all numerical methods and to provide their seamless integration. In this paper we propose the Hybrid Flow System Specification (HyFlow) as a unifying representation for different families of numerical integrators for solving ordinary differential equations (ODEs). HyFlow combines the conventional discrete event representation with a novel representation based on sampling and the support for dense outputs to describe modular and hierarchical hybrid systems. We demonstrate that HyFlow can describe 1st-order, geometric (2nd-order), and exponential integrators. Additionally, since they share the same underlaying HyFlow representation, these solvers can be seamlessly integrated.

We define an operational (transition system) semantics for the two most basic forms of Discrete Event Simulation (DES): event-based simulation (without objects) and object-event simulation. We show that under our operational semantics, DES models correspond to a certain form of abstract state machines (ASMs) such that the Future Event List (FEL) is part of the transition system state and the transition function is based on event routines. Unlike other formalisms proposed for DES (such as Petri Nets or DEVS), our ASM semantics takes all basic DES concepts (like event types and the FEL) into consideration and allows for expressive transition system states representing the objects, properties, relations and functions of the evolving possible worlds of a simulation run. As a direct formal semantics of DES, it provides a basis for comparing, and explaining design choices in, different DES approaches.

The Discrete Event System Specification (DEVS) formalism has become an engine for advances in modeling and simulation technology. Many extensions of the DEVS formalism have been developed across the years in order to solve different types of situations. However, when the acceptance of input events and the generation of output events are related to model capabilities, the current formalisms come up with complex modeling solutions. This paper presents a new simulation formalism called Routed DEVS (RDEVS) in which routing information is used to manage directed events. The behavior supported by the new formalism is useful to create simulation models of web application architectures. However, it could also be applied to other contexts. The RDEVS formalism is based on DEVS and is closure under coupling (i.e. models can be built hierarchically). The formal specification of RDEVS formalism and a briefly description of its framework implementation are presented in this work.

There has long been a divide in synchronization approaches for parallel discrete event simulation, between conservative methods requiring lookahead and optimistic methods requiring rollback. These are usually seen as dichotomous, so that a model writer must make an early, static design decision between them. An optimistic simulator does not need lookahead information but is unable to take advantage of it even if it were available, whereas a conservative simulator may perform poorly or even deadlock without good lookahead information. Here we introduce unified virtual time (UVT) synchronization which provides the advantages of both conservative and optimistic synchronization dynamically for all models. Conservative synchronization becomes an accelerator for optimistic synchronization. When lookahead information is available the simulation will execute conservatively. Otherwise it will execute optimistically. In this paper we present UVT, argue for its correctness, and show adaptations of Time Warp, YAWNS, and Null Messages which cooperatively synchronize a single simulation.

In the past few years, we have witnessed an increased interest in using multithreading PDES on multicore platforms. The work-stealing scheme, which towards to general multithread computing, can be utilized in PDES to achieve load balance straightly. However, to the best of our knowledge, the work-stealing scheme has only served as a competitor, instead of a cooperator, to other load balancing algorithm. In this paper, we propose a work-stealing based dynamic load balancing algorithm (WS-DLB) with the aim of combining their advantages. It adaptively rebalances the LPs distribution based on a priori estimation, and uses a greedy lock-free work-stealing scheme to eliminate bias at runtime. In addition, these two schemes are well adapted to enhance each other. We analyze the performance characteristics of the proposed algorithm by means of a synthetic benchmark. Experiments demonstrate that our WS-DLB algorithm achieves better performance.

Energy and power aware data distribution methods are essential for using these approaches in energy constrained devices and environments. Data Distribution Management (DDM) is a set of services defined in the High Level Architecture (HLA) that aims to efficiently propagate distributed simulation state information. This paper describes an empirical study of the energy consumption of computation and communication components of region based and grid based DDM approaches in mobile devices. Experimental data illustrate that region based approaches tend to consume more energy than grid based approaches for computations, but less for communications. These results also show that the choice of grid cell size and grid cell constraints on publication regions can play an important role in the energy efficiency of grid based approaches.

Gambling lines provide a rich source of data for Monte Carlo simulations of sports seasons. Since data for complete seasons are available after the seasons are completed, we refer to such simulations as retroactive gambling line simulations. We use the motivating and expository example of comparing the performances of the 2015-16 Golden State Warriors (who won a record 73 games) and the 1995-96 Chicago Bulls (who held the previous record of 72 wins) and provide a detailed treatment of this application. The more general applicability of the approach is examined by describing the flexibility in building the models and relating this flexibility to a few specific examples. We hope this paper will seed further explorations of retroactive gambling line simulation (and settle the Warriors vs. Bulls comparison along the way).

Captivate is a mobile game for STEM in higher education. In the development of Captivate, the employment of popular game mechanics and development of a set of mathematical tools have allowed developers to overcome some of the problems commonly facing developers of educational STEM games for higher education. Game mechanics from a variety of popular games are incorporated which supports player engagement by providing a familiar narrative to each game. Additionally, mathematical tools have been developed which provide symbolic and numeric manipulation capabilities within the game. These tools allow for graphing, problem generation and display, multiple types of player input, and evaluation of player answers all to be conducted at run time. This paper details how challenges were overcome and addressed in Captivate with examples, namely display and manipulation of complex expressions and the complexity and length of problems.

The dominance of cash-based transactions and relentless growth of a shadow economy triggered a fiscal intervention by the Indian government wherein 86% of the total cash in circulation was pulled out in a sudden announcement on November 8, 2016. This disruptive initiative resulted into prolonged cash shortages, financial inconvenience, and crisis situation to cross-section of population of the country. Overall, the initiative has faced a lot of criticism as being poorly thought through and inadequately planned. We claim that these emerging adverse conditions could have been anticipated well in advance with appropriate experimental setup. We further claim that the efficacy of possible courses of actions for managing critical situations, and probable consequences of the courses of action could have been estimated in a laboratory setting. This paper justifies our claims with an experimental setup relying on what-if analysis using an actor-based bottom up simulation approach.

Simulation studies are intricate processes that require interweaving model refinement and executing diverse experiments. Simulation models and data are the result of complex and interactive model and data generating processes. Information about these processes are required to assess the quality of simulation products. Capturing provenance, i.e., information about how a product has been generated, is a major concern both for assessing and reproducing scientific experiments. For parts of a simulation study, support for capturing and managing provenance is available. However, still gaps exist, e.g., how simulation models have been generated and to look therefore beyond individual simulation experiments and even simulation studies. To bridge those gaps it will be central to exploit, refine, and combine diverse methods effectively, as we demonstrate on a concrete case study and its provenance model.

Standards for data exchange are critical to the development of any field. They enable researchers and practitioners to exchange information reliably, to apply a variety of tools to their problems, and to reproduce scientific results. Over the past two decades, a range of standards have been developed to facilitate the exchange and reuse of information in the domain of representation and modeling of biological systems. These standards are complementary, so the interactions between their developers increased over time. By the end of the last decade, the community of researchers decided that more interoperability is required between the standards, and that common development is needed to make better use of effort, time, and money devoted to this activity. The COmputational MOdeling in Biology NEtwork (COMBINE) was created to enable the sharing of resources, tools, and other infrastructure. This paper provides a brief history of this endeavor and the challenges that remain.

In this paper, we present DiSH, a simulator for large discrete models of biological signal transduction pathways, capable of simulating networks with multi-valued elements in both deterministic and stochastic manner. The simulator incorporates the timing of molecular reactions, which are often not synchronized and occur in random order, and it also takes into account the difference between slow and fast reactions. DiSH also allows for changes in conditions during simulations, combined deterministic and stochastic changes, and it uses the concept of delays in models, similar to delays in digital circuits. DiSH is publicly available and has been used to study intra- and inter-cellular models of several diseases. It is also being used as part of a larger architecture including natural language processing tools that read biological literature, automated model assembly tools, and formal model analysis tools.

As the US Department of Energy (DOE) invests in exascale computing, performance modeling of physics codes on CPUs remains a hard challenge in computational co-design, specifically, due to the complex design of processors. Herein, we introduce Analytical Memory Model (AMM), a novel hardware model based on cache memory hierarchies that predicts the runtimes of scientific applications while addressing the co-design challenges. AMM runtime prediction depends on the distribution of reuse distances, which are measured from the memory trace of the scientific applications. The analytical reuse distribution is useful to estimate the effective latency and throughput of a memory access, which in turn are used to predict the overall runtime of a scientific application. The experimental results show that the predicted and actual runtimes of two scientific mini-applications (matrix multiplication and Blackscholes) are similar, while AMM is a scalable approach.

The paper presents a simulator designed specifically for evaluating job scheduling algorithms on large-scale HPC systems. The simulator was developed based on the Performance Prediction Toolkit (PPT), which is a parallel discrete-event simulator written in Python for rapid assessment and performance prediction of large-scale scientific applications on supercomputers. The proposed job scheduler simulator incorporates PPT's application models, and when coupled with the sufficiently detailed architecture models, can represent more realistic job runtime behaviors. Consequently, the simulator can evaluate different job scheduling and task mapping algorithms on the specific target HPC platforms more accurately.

The next generation of leadership supercomputer systems will require a medium-term layer of storage. The basis for this stratum of storage will be a Storage I/O Node (SION). For increased reliability, a redundancy algorithm will be implemented on top of groups of SIONs. In addition to the overheads of implementing a redundancy mechanism, a large cost of using a RAID strategy comes from the possibility of increased network congestion due to rebuild operations.

To better understand the impact of RAID rebuild traffic, we have developed a simulation model of the SIONs. After validation, we use this model to investigate the impact of several configuration parameters, including redundancy mechanism and the physical arrangement of hardware. Additionally, our model analyzes the use of Quality of Service lanes to limit the impact of RAID traffic. We conclude with a series of recommendations for configuring a resilient and high performing I/O subsystem.

Multi-Paradigm Modelling (MPM) has been proposed to tackle the complexities found in Cyber-Physical Systems. MPM advocates the explicit modelling of all pertinent parts and aspects of complex systems. It adresses and integrates three orthogonal dimensions: multi-abstraction modelling, concerned with the (refinement, generalization, ...) relationships between models; multi-formalism modelling, concerned with the (multi-view, multi-component, ...) coupling of and transformation between models described in different formalisms; explicitly modelling the often complex, concurrent workflows. Current modelling, analysis and simulation tools support only isolated parts of MPM.

The core methods and techniques enabling MPM are modelling language engineering, model operations, and workflow modelling. This paper delves into each enabler, presenting their relation to MPM and how they are supported in our prototype tool: the Modelverse. An automotive power window safety verification example is used to illustrate the Modelverse's capabilities. All aspects are explicitly modelled and enacted with a Formalism Transformation Graph + Process Model.

Service systems are highly dependent on staffing decisions to provide satisfactory quality of service. This paper tackles the problem of decision making under uncertainty pertaining to the source of demand. Regardless of the distribution of the demand, the proposed staffing rule reacts to the requested quality of service to determine the quality of the estimators of the unknown demand-process parameters, as well as making optimal staffing decisions. Theoretical results on the consistency and optimality of the proposed method is illustrated using sequential statistics approaches.

Analysis models, such as discrete event simulation models, are used to support design and operation of discrete event logistics systems (DELS). The time and expertise required to construct these analysis models can be significantly reduced by automatically generating them from formal models of the systems being analyzed. Many analysis models of DELS are constructed from system abstractions, but when the system and analysis are specified at incompatible levels of abstraction, the relationship between the two is imprecise and ad-hoc. Formal modeling languages, such as those used in object-orientation, make abstraction explicit, simplifying the mappings between system and analysis models and increasing reusability of the integration. We propose fundamental abstractions for DELS and identify corresponding libraries of analysis models. These are used in a system-analysis integration methodology that incorporates abstraction as an explicit step, providing a path to refine those abstractions and model libraries and to generate analysis models from them.

The size and complexity of modern power systems, as well as emergent technologies and the uncertainty in energy planning, make the design and engineering of these systems challenging. One of the main challenges is the development of models that adequately capture the complex relationships between the components of these systems. We present a DEVS (Discrete Event System Specification) based framework for modeling a power system for energy planning. DEVS preserves the hierarchical and modular construction properties of a system (Chow and Zeigler 1994). That is, it enables each of the components of the system to be modeled separately, as well as the representation of the multilayer architecture of that system. As a proof of concept, we present a power system model, simulating the deployment of energy sources, on the PyPDEVS platform (Van Tendeloo and Vangheluwe 2015), considering dispatcher, unit commitment, load, generation, storage and transmission lines components.

We present a cellular model to study the propagation of cracks in a given sample of rock. The model uses the fractal geometry found in real rocks and the stress and strain on an element of rock to update the element’s strength. We study the propagation speed of the cracks with different initial conditions. The model shows that the more inhomogeneous a rock sample is the quicker a fracture can propagate through it. Because fractures release a large amount of energy when they grow, earthquakes are more likely to occur in these situations.

Markov decision processes (MDPs) have proven useful as models of stochastic planning and decision problems. To try to propose practical implementation of MDPs, hierarchical methods are often used in MDPs or reinforcement learning to delegate the optimization of the total problem to simpler hierarchical subproblems. The goal of the paper is to propose a generic discrete-event based software Framework allowing to use hierarchical MDPs and reinforcement Learning to solve planning or decision problems. The proposed approach has been validated using the ”grid world” typical MDP use case.

Air traffic management is widely studied in several different fields because of its complexity and criticality to a variety of stakeholders. However, the exploding amount of air traffic in recent years has created new challenges to ensure effective management of the airspace. A fast time simulation capability is essential to effectively explore the consequences of decisions of air traffic management. A new algorithm for simulating air traffic networks using a time-parallel simulation approach is proposed that distributes time segments of the simulation scenarios across different processors. A simulation model for the National Airspace System (NAS) is described and validated. The components of the simulator are described as well as the parallel simulation algorithms. Experimental results utilizing real-world traffic data for the continental U.S. are presented demonstrating the speed ups achieved by a prototype simulator. These results illustrate that time-parallel simulation can be used to significantly accelerate certain air traffic simulations.

This paper summarizes the profile data captured from 22 discrete-event simulation models captured from 4 different simulators. The profile data is captured by instrumenting the simulation engine and does not require any modification to the models. The profile data reported focuses on the communication properties of events exchanged between the various processes (LPs) in the model. The data suggests that some models share common behaviors and this work summaries their general characteristics. This permits a presentation of the principle characteristics using only six of the studied models. This resulting information can be used to: (i) provide configuration data for synthetic model generation, (ii) provide direction for configuring and optimizing parallel and distributed simulation engines, and (iii) provide insights into model correctness. The focus of this specific study is to determine regularities among LP event communications that can be exploited for model partitioning and event scheduling in parallel simulation.

The Naming Game is an effective self-organization model to explain and understand the emergence of linguistic consensus and system dynamics in a variety of phenomena over social networks of autonomous agents. It is an effective description for the evolution of consensus despite the absence of any central coordination or specialized initialization even in large-scale networks. While the classical game is effective in description, it was defined with inherently sequential evaluation semantics over the entire network. Here, we develop a new concurrent model as a relaxation of the classical Naming Game and express the concurrent model in a discrete event style. Further, with the uncovered concurrency absent in the classical algorithm, we map the concurrent model to parallel discrete event simulation. Using a prototype implementation, we present an initial parallel simulation performance study on networks containing hundreds of thousands of individuals, with simulation time decreased from 4800 seconds to 1400 seconds.

Simulation is an important tool to validate the performance impact of control decisions in Software Defined Networks (SDN). Yet, the manual modeling of complex topologies that may change often during a design process can be a tedious error-prone task. We present TopoGen, a general purpose architecture and tool for systematic translation and generation of network topologies. TopoGen can be used to generate network simulation models automatically by querying information available at diverse sources, notably SDN controllers. The DEVS modeling and simulation framework facilitates a systematic translation of structured knowledge about a network topology into a formal modular and hierarchical coupling of preexisting or new models of network entities (physical or logical). TopoGen can be flexibly extended with new parsers and generators to grow its scope of applicability. This allows to design arbitrary workflows of topology transformations. We tested TopoGen in a network engineering project for the ATLAS detector at CERN.

As smart grid systems become increasingly reliant on networks of control devices, attacks on their inherent security vulnerabilities could lead to catastrophic failures. Network Intrusion Detection Systems(NIDS) detect attacks by learning traffic patterns and finding anomalies in them. However, availability of data for robust training and evaluation of NIDS is rare due to associated operational and security risks of sharing such data. Consequently, we present Melody, a scalable framework for synthesizing such datasets. Melody models both, cyber and physical components of the smart grid by integrating a simulated physical network with an emulated cyber network while using virtual time for high temporal fidelity. We present a systematic approach to generate traffic representing multi-stage attacks, where each stage is either emulated or recreated by replaying arbitrary packet traces. We evaluate the suitability of Melody’s datasets for intrusion detection, by analyzing the extent to which temporal accuracy of pertinent features is maintained.

Driven by the needs to monitor, detect, and prevent catastrophic failures for complex information systems in real-time, we develop in this paper a discrete-time queuing network simulator. The dynamic model for the simulator abstracts network dynamics by taking an aggregated and layered structure. Comparative studies verify capabilities of the simulator in terms of accuracy and computational efficiency. We illustrate the model structure, flow processing mechanisms, and simulator implementation. We also illustrate the use of this simulator to detect distributed denial-of-service (DDoS) flooding attacks, based on a cross-correlation-based measure. Finally, we show that the layered structure provides new insights on the spatiotemporal spread patterns of cascading failure, by revealing spreads both horizontally within a sub-network and vertically across sub-networks.

Vector-borne infectious diseases present computational modelers with a unique challenge: finding the right balance between model fidelity and simulation costs. In this work, we introduce a hybrid agent-based model that achieves a balance that readily scales to problem sizes of millions of human agents and mosquitoes. Macroscopically, our model results agree with those from a low-cost compartmental model; microscopically, like agent-based models, it provides details at the individual level. We apply this model to a synthetic human population of 1.2 million individuals from Miami, Florida in the United States to model the Zika outbreak in the Fall of 2016. We identify two high-risk locations within this region, a detail which cannot be revealed by traditional compartmental models. The principles-based mathematical derivation of the hybrid agent-based model can be adapted to other scenarios facing similar tradeoffs.

Many high-burden countries have committed to providing universal access to rapid diagnosis of tuberculosis (TB), but the corresponding impact on population-wide incidence is unknown. We designed an agent-based simulation of drug-susceptible (DS) and drug–resistant (DR) TB in a representative Indian setting and compared the impact of Xpert testing via a decentralized (Xpert available at each local-population) versus centralized (Xpert available at the district-level serving multiple local-populations) strategy. Decentralized testing resulted in a 36% reduction in DR-TB incidence at 10 years compared to no Xpert. Depending on assumptions regarding pre-treatment loss to follow-up (ranging from 5 to 50%), the impact of centralized testing ranged from a 35% to 22% reduction in DR-TB incidence. Implementation of Xpert by either approach had a negligible impact (<5%) on DS-TB incidence. Decisions regarding choice of centralized vs. decentralized Xpert will heavily depend on operational aspects of centralized Xpert and loss to follow-up.

The objective of this study was to specify and model the behavior of households regarding the installation of water conservation technology and evaluate strategies that could potentially increase water conservation technology adoption at the household level. In particular, this study created an agent-based modeling framework in order to understand various factors and dynamic behaviors affecting the adoption of water conservation technology by households. The model captures various demographic characteristics, household attributes, social network influence, and pricing policies; and then evaluates their influence simultaneously on household decisions in the adoption of water conservation technology. The application of the proposed simulation model was demonstrated in a case study of the City of Miami Beach. The simulation results identified the intersectional effects of various factors in household water conservation technology adoption and also investigated the scenario landscape of the adoptions to inform policy formulation and planning.

This article aims to respond to growing concerns about sustainable urbanization, which in recent years have generated a need for prospective assessment in the field of transport and land-use planning, by predicting future land-use development. We introduce a Land Use model (part LU of a Land Use Transport Interaction model) which aims to simulate households and firms location choice within an urban system.. We use the agent-based approach to simulate location choices in order to account for land use changes and to estimate residential and economic activities location. This is a dynamic bottom-up approach with the households and the firms as their basic components. The MUST-B model considers the agents’ location choices according to the utility theory and the equilibrium between real estate supply and demand. The model is used to simulate urban land-use development in the urban area of Bordeaux, France.

Photovoltaic panels generate electricity directly from sunlight, making them a favored renewable technology. Green roofs are rooftops covered with vegetation, which provide a variety of benefits, namely, reducing stormwater runoff, improving air quality, and biodiversity. Green roofs are capable of improving the efficiency of Photovoltaic panels, as shown by the recent studies. Optimal placement of Photovoltaic panels and green roofs is a challenging problem due to the complications imposed by uncertainties associated with future climate conditions, specifically due to climate change. An agent based model to optimally place Photovoltaic panels and green roofs is developed in this study. We propose a tabu search metaheuristic algorithm to solve the developed model. Then, a real-world case for a mid-sized city in the U.S. is solved as a case study for the model. We further conduct numerical analysis and provide insights.

A building occupancy simulation and estimation simulates the dynamics of occupants and estimates the real time spatial distribution of occupants in a building. It needs a simulation model and a data assimilation algorithm that assimilates real-time sensor data into the simulation model. In our previous works, we presented a graph-based agent-oriented simulation model, and a data assimilation framework based on Sequential Monte Carlo (SMC) methods for efficient real time occupancy estimation involving a large number of occupants. As the occupancy and the building environment size increases, there is a major problem with data assimilation caused by the high dimensional system states. To address this issue, this paper presents a new sensor-informed resampling method which utilizes sensor data to improve resampling of particles. Experiment results show the new method was able to provide better estimation of the system state with a limited number of particles compared to the standard bootstrap filter.

In this paper, we introduce the agent-based simulation of a shopping mall with walking and purchasing behavior model and consider the performance of distributed parallel execution. To utilize the agent-based simulation for decision support, distributed parallel execution of large-scale agent-based social simulations is important for evaluating the complex behavior of a realistic number of people with acceptable performance. For this purpose, today's agent-based simulation frameworks often provide the functionality to transfer agents from one node to another. However, intelligent social agents tend to contain a large amount of data including demographics, preferences, and history. Hence, the transfer of such an agent incurs a heavy communication cost that has an adverse effect on performance. To improve the performance of distributed agent-based simulation, we introduce a shadow agent that is a lightweight entity projected among nodes with only required information such as the position and speed required to calculate interaction between agents.

Crowd simulation is often used as a crucial tool to analyse crowd behaviours. Ideally, when analysing live video streams, we would like the simulator to be able to run concurrently. However, crowd video analytics algorithms are usually not able to supply position updates in real time as there exists a noticeable time gap between two consecutive human position updates. the crucial problem is therefore on how to simulate human positions within the time gap. In this paper, a simulation framework that could approximate human displacements in a near real time manner is proposed. A framework based on OpenCV that reads video streams and runs real time simulation is implemented. As a result, amongst the crowd being tracked, we obtain near real time simulation with acceptable tracking accuracy. Lastly, this paper explains the limitation of the proposed framework.

Both the scientific community and the popular press have paid much attention to the speed of the Securities Information Processor – the data feed consolidating all trades and quotes across the US stock market. Rather than the speed of the Securities Information Processor, or SIP, we focus here on its importance to efficient, price discovery. Via extensions to a previous market model, we experiment with four different coupling mechanisms which operate across the US stock market. Of the four, we find that the SIP contributes most to efficient price discovery.

We describe an agent-based stock market simulator built using an asynchronous discrete event simulation framework. The simulator is unique in that it’s driven by real-world financial algorithms and protocols; and it’s open source. It utilizes an order book bid and ask matching model, and real-world exchange protocols. Our simulation is based on multiple agents interacting through an exchange agent. This method is distinct from those that utilize historical pricing data. Order book execution supports a more realistic interaction between agents. Pricing in our model arises from the dynamics of matching orders in the order book. Our simulator enables the study of market dynamics, and trading strategies using real-world exchange protocols. We present our design and implementation of a market simulator and discuss our initial results using the message protocols defined by NASDAQ: OUCH and ITCH. Our initial results demonstrate StockYard’s capability and efficiency in simulating markets with realistic trade volumes.

In this paper, an agent based model of stock market is developed. All agents fall under three general trading systems: fundamentalists, optimists and pessimists, while having different beliefs within each category. The model stresses interaction and learning among adaptive agents, which causes macroscopic properties to emerge in the market. The generated time series revealed that the model was able to reproduce some of the key stylized facts observed in actual financial time series and was consistent with empirical observations. Herding is one of the important emergent properties of financial markets, often leading to the creation of speculative bubbles that make markets unstable and prone to crashes. Using this validated framework, the necessary infrastructure is provided to explore the relationships of different market properties with this phenomenon. The study suggests that herding has significant causal relationship with volatility in the market, and vice versa.

Surrogate models are commonly used to approximate the multivariate input or output behavior of complex systems. In this paper, surrogate assisted calibration frameworks are proposed to calibrate the crowd model. To integrate the surrogate models into the evolutionary calibration framework, both the offline and online training based approaches are developed. The offline training needs to generate training set in advance, while the online training can adaptively build and re-build the surrogate model along the evolutionary process. Our simulation results demonstrate that the surrogate assisted calibration framework with the online training is effective and the surrogate model using artificial neural network obtains the best overall performance in the scenario evaluated in the case study.

Network-based attacks like the distributed denial-of-service (DDoS) attacks are not new, but we are beginning to see attacks of unprecedented scale. Examples of such attacks include the 2016 attack on DYN INC that crippled a part of the Internet for hours, and the attack on Liberia, which partially brought down the African nation. Limitations in identifying vulnerable Internet infrastructure and testing possible defense strategies are a part of the problem. We need a simulation testbed that can reflect the complexity of the Internet, yet allows to swiftly test attacks, providing insights that can apply to real-world attack scenarios. In this research, we have designed a test-bed that mirrors the Internet infrastructure of the US and can simulate the Internet traffic flow patterns for different attack targets. We also estimate the degradation in the quality-of-service and the number of users impacted in two attack scenarios.

Forced migration is a growing global problem, and the world now has a record amount of 22.5 million refugees. Models that predict refugee movements are few and far between, and constructing these models requires a substantial amount of manual effort while erupting refugee crises require a very rapid response. Here we present a vision towards establishing an automated framework, aimed to enable researchers to construct simulations of refugee movements more quickly and systematically. Our approach incorporates a diverse range of data sources, and uses the FabSim toolkit in conjunction with the Flee simulation code to quickly generate simulation workflows. In addition, we highlight a few key steps that we have already taken towards realizing this vision and discuss opportunities for wider applicability.

The intracellular calcium signaling pathways of a neuron consist of biochemical reactions along with molecular diffusion. Stochastic discrete event simulation of these pathways provides a more detailed understanding of the pathways than deterministic simulators because they capture behavior at a molecular level. Our research employs a parallel discrete event simulation simulator, Neuron Time Warp (NTW), which is intended for use for the simulation of neurons. We did not achieve the expected performance of previously derived Ca^(2+) wave model because of an imbalance in computation between the area of the neuron covered by the Ca^(2+) wave and the remaining area of the neuron. In this paper we describe a dynamic load balancing algorithm with a dynamic window control algorithm for NTW. We make use of Q-learning to determine the basic parameters of the algorithm. Using this algorithm we obtained an improvement in the performance of the simulator of up to 30%.

Agent-based modeling (ABM) assumes that behavioral rules affecting an agent’s states and actions are known. However, discovering these rules is often challenging and requires deep insight about an agent’s behaviors. Inverse reinforcement learning (IRL) can complement ABM by providing a systematic way to find behavioral rules from data. IRL frames learning behavioral rules as a problem of recovering motivations from observed behavior and generating rules consistent with these motivations. In this paper, we propose a method to construct an agent-based model directly from data using IRL. We explain each step of the proposed method and describe challenges that may occur during implementation. Our experimental results show that the proposed method can extract rules and construct an agent-based model with rich but concise behavioral rules for agents while still maintaining aggregate-level properties.

In an agent-based simulation, a plausible decision in a specific context cannot stay be valid in the face of the changing situation. Therefore, the result of the decision making process is mostly related to the agent’s situational awareness and its adaptation to the new context and situation of the environment. A sound estimation of the situation requires a clean understanding of the operational domain, not only the data taken from sensors. In this paper, it is aimed to improve decision making by increasing the situational awareness of an agent by incorporating the decision making mechanism with the prior knowledge about the problem domain, such as the existing rules. Specifically, an existing adaptive decision making architecture, which is based on the deliberative coherence theory, is adopted to be driven by situational awareness. Moreover, a case study incorporating unmanned surface vehicles that are aware of the international sea traffic rules is demonstrated.

The use of simulation modeling for scientific tasks demands that these models be replicated and independently verified by other members of the scientific community. However, determining whether two independently developed simulation models are “equal,” is not trivial. Model alignment is the term for this type of comparison. In this paper, we present an extension of the model alignment methodology for comparing the outcome of two simulation models that searches the response surface of both models for significant differences. Our approach incorporates elements of both optimization and design of experiments for achieving this goal. We discuss the general framework of our methodology, its feasibility of implementation, as well as some of the obstacles we foresee in its generalized application.

There are often concerns about the reliability of simulation results due to improper design of experiments, limited support in the execution and analysis of experiments, and lack of integrated computational frameworks for model learning through simulation experiments. Such issues result in flawed analysis as well as misdirected human and computational effort. We put forward a methodological basis, which aims to (1) explore the utility of viewing models as adaptive agents that mediate among domain theories, data, requirements, principles, and analogies, (2) underline the role of cognitive assistance for model discovery, experimentation, and evidence evaluation so as to differentiate between competing models and to attain a balance between model exploration and exploitation, and (3) examine strategies for explanatory justification of model assumptions via cognitive models that explicate coherence judgements.

Simulation experiments contribute to scientific discovery due to the degree and extent of reproducibility that simulation systems provide. On the other hand, domain scientists may lack expertise in simulation programming and the use of effective methods for instrumenting, evaluating, and comparing models. By utilizing formal automated verification methods, we aim to improve the process of evaluating model assumptions against evidence, and to facilitate selection of new hypotheses to maximize information gain while reducing information processing requirements. To this end, to evaluate the results of a simulation experiment against expected regularities, a probabilistic model checking system is coupled with a Domain-Specific Language that expresses abstract finite state verification properties. These specification patterns are evaluated against the run-time Discrete-Time-Markov Chain model abstracted from the data obtained through aspect-driven automated instrumentation.

This paper presents a simulation model of a generic automated planning and control system for the pick-up and docking of semi-trailers by means of autonomous Yard Tractors (YTs) in a collision- and conflict free environment. To support the planning and control of the YTs, we propose a Multi-Agent System (MAS). We illustrate our approach using a case study at a Dutch logistics service provider. To evaluate the proposed system, we design an agent-based simulation model, which is set up in a similar way as the MAS. We conclude with the verification and validation of the simulation model.

Vehicle sharing can increase the efficiency of transportation infrastructures and improve environmental sustainability. A distributed operation model is needed to improve a vehicle’s intelligence and autonomy. In this research, we develop an agent-based simulation model for a linear transportation system to evaluate three different vehicle sharing operations that include: 1) an independent operation where vehicles are isolate, 2) a centralized operation which assumes a central supervisor agent controls all the vehicles, and 3) a distributed operation where vehicles can communicate with others and make decisions by themselves. Our simulation results demonstrate that: 1) the centralized and distributed models are significantly better than the independent model for large car capacity, 2) centralized model performs significantly better than the distributed model for large car capacity and small communication range, and 3) the distributed model can perform better than the centralized model for large car capacity and communication range.

By explicitly modeling the decision making of heterogeneous individuals, agent-based models can compute the resulting emergent phenomena on the micro-level. This lets planners evaluate new planning approaches for problems that depend on individual decisions. Examples include airline revenue management or traffic control. However, when decision support relies on agent-based modeling, its applicability to real-world problems depends on the model's validity. This paper introduces a novel methodological concept to decompose agent-based models for calibration and validation. This concept enables modelers to isolate agents from the evolution of the model's state variables, allowing greater choice of calibration and validation approaches. The approach first parameterizes and validates individual agents, and subsequently re-calibrates the agent-collective within the entire model.

In simulation studies, the goal specifies the objective or purpose of the study and thus drives the entire experimentation process. Relevant experiments and respective experiment hypotheses are derived from the study's goal and the model's observed behavior provides evidence whether these hypotheses hold. Current assistance systems do not integrate research hypotheses. Thus, the researcher has to make important design decisions which limits both replicability and reproducibility of the results. In this paper, the process of simulation studies is systematized based on a formally specified hypothesis. By this means, the research hypothesis becomes the key element of the study and guides the entire process. Hypothesis-driven simulation studies allow for the automated design, execution, and evaluation of experiments based on specific research questions. This facilitates documentation and execution as well as replication of simulation studies.

Trustworthy statistical modeling is an emerging challenge in agent-based modeling (ABM). However, typical characteristics of ABM such as the potential for high numbers of entities and parameters, interdependent relations between entities, several layers of effects, and emergent social phenomena challenge this process. In particular, aggregated outcomes emerging from individual agent interactions are, at least partly, difficult to measure. This might impede the statistical modeling process and thus the formulation of trustable conclusions. For this reason, we introduce structural equation modeling (SEM) as a promising statistical modeling method to analyze the behavior of ABMs. SEM allows for the estimation and evaluation of highly networked systems by explicating interactions between types of agents, measuring emergent phenomena, and identifying output patterns of simulation models. Overall, these contributions foster the credibility and trustworthiness of ABMs, and also support the communication and understanding of simulation models’ behavior and their output.

Coordinated search processes are pervasive in both organizations and product development projects. In such processes, designers with different specialties learn about their interdependent alternatives through a mutual adjustment process. In the context of a product development with several teams developing the new product's subsystems, and using reinforcement learning and agent-based simulation modeling, this study looks at the performance effects of design teams' initial mental characterizations about subsystem interactions. The focus is on two initial mental models, one in which teams over-weight their own subsystem's element interactions, and another, in which teams over-weighting interactions between subsystems. The results indicate that both initial representations have critical performance consequences for product development. Specifically, teams prioritizing their interactions of their own subsystem's elements gain short-run performance benefits as they converge to a local optimum in a short time period. Contrarily, over-weighting between-subsystem interactions leads to a tendency for teams to have long-run performance advantages.

This paper studies social influence (i.e., adoption of belief) using agent-based simulation and regression models. Each agent is modeled by a linear regression model. Agents interact with neighbors by exchanging social beliefs. It is observed that if individual belief is linear in neighbors’ beliefs, system-level belief and aggregated neighbors’ beliefs can also be described by a linear regression model. Analysis is conducted on a simplified 2-node network to provide insight into the interactions and results of general models. Least squares estimates are developed. Explicit expressions are obtained to explain relationship between initial belief and current belief.

This paper introduces a new consumer decision-making model where each agent uses a neural network to evaluate word-of-mouth and predict her utility prior to adoption a new product based on her experiences in the past. The model considers the fact that consumers may not know their true preferences before experiencing the product. By using a neural network, an agent can: (1) interpret the feedback from a neighbor who has conflicting preferences with her; (2) interpret partially positive and/or negative feedback; and, (3) assign different weights to the feedback received from different neighbors. The model is implemented in an agent-based simulation model to verify that the resulting diffusion dynamics follow a typical diffusion curve. Preliminary experiments with the model also provide interesting results about the effect of the number of product attributes on the quality of an individual's utility prediction as well as proportion of satisfied adopters.

Probabilistic timed automata are a formal model for real-time systems with discrete probabilistic and nondeterministic choices. To overcome the state space explosion problem of exhaustive verification, a symbolic simulation-based approach that soundly treats nondeterminism to approximate maximum and minimum reachability probabilities has recently become available. Its use of difference-bound matrices to handle continuous real time however leads to poor performance: most operations are cubic or even exponential in the number of clock variables. In this paper, we propose a novel region-based approach and data structure that reduce the complexity of all operations to being linear. It relies on a particular mapping between symbolic regions and concrete representative valuations. Using an implementation within the Modest Toolset, we show that the new approach is not only easier to implement, but indeed significantly outperforms all current alternatives on standard benchmark models.

Verifying software in mission-critical Cyber-Physical Systems (CPS) is an important but daunting task, faced with challenges of accuracy and scalability.

This paper discusses lessons learned from verifying properties of the Linux kernel. These lessons have raised questions about traditional verification approaches, and have led us to a model-based approach for software verification. These models are high-level models of the software, as opposed to the prevalent formal methods with low-level representations of software.

We use models to gain insights into the challenges of software verification and use those insights to improve software verification. We demonstrate significant advantages of models with a Linux kernel study involving verification of 66,609 Lock instances. We use models to: (a) analyze and find flaws in verification results from LDV, a top-rated Linux verification tool, (b) show significant improvement over LDV by improving accuracy, speed, and by verifying 99.3% instances compared to 65.7% instances by LDV.

In this paper, we consider a futuristic scenario where there exists a special lane in a segment of a highway; vehicles which wish to use the lane must send access requests ahead of time; and only the vehicles whose requests are accepted can use the lane. Vehicle classes are defined by vehicle sizes and carry a different number of passengers. Our goal is to find an optimal allocation of the lane capacity among different vehicle classes maximizing the long-run average passenger throughput. We use a multi-class M/G/C/C state dependent model to calculate the long-run average passenger throughput for a given allocation and use a cross-entropy method to find an optimal allocation among a large number of possible allocations. We discuss how this problem formulation is applicable to general resource sharing problems and discuss how to dynamically control acceptance/rejection of access requests to further enhance the real-time efficiency of our system.

Intelligent transportation systems are typical Cyber-Physical Systems (CPS) that combine physical components with cyber elements that include communication, information processing and control mechanisms for Connected Automated Vehicles (CAVs). To test and evaluate the efficiency of such systems, new simulation platforms are needed. In this paper, a SimEvents-based framework is introduced for hybrid traffic simulation at the microscopic level. This framework enables users to apply different control strategies for CAVs and carry out performance analysis of proposed algorithms by authoring customized discrete-event and hybrid systems based on MATLAB Discrete-Event System using object-oriented MATLAB. The framework spans multiple toolboxes including MATLAB, Simulink, and SimEvents.

The research and development of novel user experience concepts in well-regulated industrial domains face different challenges. Systems in these domains often require backward compatibility and integration with legacy sub-systems and protocols. They must comply with well-defined procedures and standards, and must pass through stringent evaluation processes involving actual users under realistic conditions and scenarios. As a consequence, prototyping and simulations are extensively used. During product development, the level of fidelity of a simulation prototype will directly impact the quality of end-user feedback, minimizing expensive rework of UX in later stages of a project. This paper describes the Distributed Industrial Simulation Platform (DISP), a simulation framework developed within GE that facilitates the rapid prototyping and evaluation of novel industrial UX systems. We present the DISP design and main services showing how it has been used in support of the development and simulation of two UX prototypes in the railroad transportation domain.

The number of projects relying on volunteer computing and their complexity are growing fast. This distributed paradigm enables the gathering of idle resources (processing power and storage) to run large systems by providing scalable, practical and low cost platforms. The heterogeneity of the resources and their unreliable behavior call for advanced optimization methods. In particular, an efficient resource allocation is key for the systems’ performance. This work presents a mathematical formulation and a solving approach based on a metaheuristic for the resource allocation problem. This approach is designed to deal with data-intensive applications, which must guarantee the availability of the data at all times. Moreover, a simheuristic is proposed to deal with the stochasticity of resources’ quality. A set of computational experiments are performed to: (1) compare the performance of the metaheuristic and the simheuristic in a stochastic environment; and (2) quantify the effect of the stochasticity on the solutions.

Software-induced delays have a significant impact on real-time control system performance. While in extreme cases, jitter caused by execution time variations or start-time delays may compromise stability, such effects are often ignored in model-based approaches and simulations. We address this issue and discuss a Simulink block with a fundamentally new semantics that is used to encapsulate control functions as software tasks. It is particularly suited to model and observe software delays in complex control functions within software-in-the-loop (SIL) simulations. For this purpose, the controller source code is associated with platform-specific execution time information, which we assume to be available. We detail a synchronization mechanism between Simulink and such time-annotated task blocks implementing the control laws. It allows us to perform SIL simulations where controllers execute for a finite amount of simulation time and span multiple simulation steps at which they may interact with other controllers or the plant.

Today's advanced machine tool and manufacturing system engineering uses mechatronic, system-based modular kits in order to offer machines and manufacturing systems economically. The market increasingly requires not only economical production of the machine tool or manufacturing system, but also the latest technology, for example, an upgrade or exchange of mechatronic modules. This is not economically possible due to the state of the art. This paper will present how mechatronic engineering can be developed into an engineering using cyber-physical systems (CPS). It will also present, how the engineering of machine tools and manufacturing systems will change in the future und which opportunities can be realized.

Modeling and simulation (M&S) is a discipline oriented towards engineering and research, but it tends since the very last years to be used more and more by users and developers of mobile applications through cloud computing and web services. The M&S new tools involve mobile terminals (smartphone, tablet, etc.) exchanging data quantities increasingly important from sensors with an increasing transmission speed. This paper presents a generic approach (the DEVSimPy-mob mobile application) which aims to simulate models described with the DEVS formalism (Discrete EVent system Specification). DEVSimPy-mob communicates with a web REST (Representational State Transfer) server that delivers a set of web services dedicated to the simulation of DEVS models. A real case application stemming from Balistic Missile Defense simulations is presented to show how DEVSimPy-mob can be used to launch simulations from a mobile device, interact during the simulation process and visualize results.

This paper presents a hybrid Modeling and Simulation framework to address business process challenges. The framework has integrated process mining techniques in the conceptual modeling phase to support developing simulation models that are unbiased and close reflection of reality in a timely manner. The hybrid approach overcomes the pitfalls of traditional conceptual modeling by using process mining techniques to discover valuable process knowledge from the analysis of event logs. The proposed hybrid framework has been applied to an Emergency Department (ED) in order to identify performance bottlenecks and explore improvement strategies in an attempt to meet national performance targets. A large number of unique process flows (i.e. patient pathways) within the ED were uncovered and deviations from process guidelines were accurately identified. Results show that unblocking of ED outflows have a significant impact on patients length of stay (over 80% improvement) rather than increasing the ED physical capacity.

Machine Learning (ML) has demonstrated great potentials for constructing new knowledge, or improving already established knowledge. Reflecting this trend, the paper lends support to the discussion of why and how should ML support the practice of modeling and simulation? Subsequently, the study goes through a use case in relation to healthcare, which aims to provide a practical perspective for integrating simulation models with data-driven insights learned by ML models. Through a realistic scenario, we utilise ML clustering in order to learn about the system’s structure and behaviour under study. The insights gained by the clustering model are then utilised to build a System Dynamics model. Recognizing its current limitations, the study is believed to serve as a kernel towards promoting further integration between simulation modeling and ML.

In face of high urbanization and increasing mobility, models and simulations are used to find answers for urban planning problems. However, simulations face criticism for over-simplifying complex reality, having models disconnected from the context of their use or excluding policy-makers from the building of models. Smart city approaches did not overcome that reality even if they relied more and more on microscopic models, together with data available through technology. This article describes a hybrid approach combining the expert knowledge on the city and its limits in terms of data, with models having the right dimensionality to provide policy-makers and urban managers with the necessary information for understanding and managing the city. This approach has been applied in Venice, but it describes in more general terms a way of bridging the world of theoretically sound models with their potential use.

The container flow in terminals at a port is often bottlenecked due to the slow operations of the quay cranes in a scarce terminal land space. The quay crane scheduling problem (QCSP) is a major problem because of the assignment of expensive quay cranes to multiple vessel-holds for container discharging and loading operation. This paper presents a hybrid QCSP Solver, which combines genetic algorithms for global search with steepest ascent hill climbing for local search. Numerical experiments are performed with small- and large-sized random QCSP instances. The experimental results revealed that the hybrid QCSP Solver provides a better solution than the stand-alone QCSP Solver. By scheduling the dynamic operation of quay cranes it is expected that the developed decision making tool will provide terminal planners with a guideline to enhancing the assignment of quay cranes to a vessel.

While simulation has been used extensively to model supply chain processes, the use of a Bayesian approach has been limited. However, Bayesian modeling brings key advantages, especially in cases of uncertainty. In this paper, we develop a data informatics model that could be used to realize a digital synchronized supply chain. To realize this model, we take a hybrid approach that combines Bayesian modeling with discrete-event simulation and apply it to the supply chain process of a Procter & Gamble (P&G) manufacturing and distribution facility. We use approximately one year of transactional data to inform our model, including information on customer orders, production, raw materials, inventory, and shipments. A driving force for creating this model is to better understand and manage the balance between inventory, profit, and service.

Healthcare services worldwide are prioritizing efficiency of delivery and optimization of resource allocation. Efficient healthcare delivery relies on the coordination of demand and capacity, but forecasting studies often predict demand without regard for future capacity constraints. Likewise, capacity planning requires strategic decision-making, therefore planning tools should allow decision-makers to examine the consequences of changing demand and likely capacities required over time. The aim of the study is to evaluate a hybrid methodology using discrete-event simulation and demand forecasting in the healthcare domain. A case-study investigates the application of official population projections with local historical demand data to forecast demand for a healthcare diagnostic service. The resultant forecasts are then used with DES in a hybrid systems modeling approach. This provides plausible demand forecasts for future capacity planning and resource allocation in a preventative healthcare service. It also contributes to debates on the value of hybrid approaches in supporting real-world decision-making.

Mixed type random variables contain both continuous and discrete components, and their role is critical in many well-studied fields. Queuing analysis, stock options, and hydrology rainfall models are among those dependent on mixed random variables to simulate event outcomes. In each of these examples, continuous positive distributions combine with a discrete spike at zero to adequately represent system uncertainty. These problems often require simulation because analytic solutions using these hybrid distributions quickly grow in complexity. Concessions are made, however, when using simulation. In addition to inherent sampling variability, perspective of discrete and continuous components is easily lost when plotting results. This paper details these challenges, and touches on the shifting line between simulations and attainable analytic results. It discusses computational probability's potential to improve model realism and accuracy, introducing MixedAPPL software prototype, an extension of Maplesoft based APPL (A Probability Programming Language) capable of manipulating mixed type random variables.

As hybrid simulation development context is a new topic with limited available applications and data, and thus still mistrusts most researches, this work proposes a methodology to combine discrete event simulation (DES) and System Dynamics (SD) paradigms on a logistics background. On the basis of knowledge induced from literature, a generic conceptual framework for hybrid simulation has been developed. The proposed framework is demonstrated using an explanatory case study comprising an user transportation mode choice. The methodology is able to create an feedback dynamic between both paradigms, ensuring advantage to hybrid methodology modeling, that is able to combine and extract the benefits from both paradigms.

Companies in a variety of industries rely on their employees to work together effectively in teams to achieve their objectives. However, finding ways to encourage collaborative behavior to optimize a team’s performance is often challenging. In particular, managers would like to be able to increase the likelihood that team members decide to help each other, in the event of workload imbalances. In this paper, a hybrid simulation (ABM-DES) model has been developed to investigate how workers’ predisposition to altruistic tendencies, an important personality factor, influences their willingness to help their co-workers on a production task. Model inputs were derived from real time laboratory experiments from which data on participants’ personalities, perceptions, and decisions to help team members complete a task were captured. Simulation results suggest that the individuals with high altruism characteristics are more likely to help their co-workers.

There is a growing trend in the number of M&S studies that report on the use of Hybrid Simulation. However, the meaning and the usage of the term varies considerably. Indeed, the hybrid simulation panel during last year’s conference (WSC2016) laid bare the strong views, from the panelists and audience alike, as to what constitutes a hybrid model and what is new? The ensuing debate set the scene for this year’s paper, in which we discuss the various perspectives on hybrid simulation by focusing on three aspects: its definition, its purpose and its benefits. We hope this paper will pave the way for further studies on this subject, with the objective of achieving a convergence of the definition of hybrid simulation.

This study has a multimethodological approach where discrete event simulation (DES) modelling and Soft Systems Methodology (SSM) is used in combination at the surgical unit of The Norwegian Radium Hospital. The aim was to investigate the effect of different interventions on patient flow and resource utilization. The multimethodological approach ensured well anchoring and a feeling of ownership of the project among all staff groups. This increases the probability of acceptance of model results. The multimethodology also ensured that it was the most relevant, desirable and feasible interventions that was evaluated with the simulation model. This project has provided the hospital management with a massive amount of structured information, based both on results from the SSM and the DES processes. Several assumptions have been quantified and are used as tools in the discussion on how to meet the increased demand of cancer surgeries due to recently implemented cancer pathways in Norway.

Home hospital services; provide some hospital level services at the patient’s residence. The services include for example : palliative care, administering chemotherapy drugs, changing dressings and care for newborns. The rationale of the service is that by providing high quality care to patients at their homes their experience of the care is better and hence they respond to the treatment and/or recover quicker and are less likely to need to report to hospital to receive care for more serious/expensive conditions. The aim of this study is to evaluate the effectiveness of the home hospital service, to optimize the current configuration given existing constraints and to evaluate potential future scenarios. Using a combined discrete event simulation, agent based model and geographical information system we assess the system effects of different demand patterns, appointment scheduling algorithms (e.g. travelling salesman problem), varying levels of resource on patient outcomes and impact on hospital visits.

Incidence rates in simulation models are often assumed to stem from Poisson processes, with rates based on analyses of real-life data. In cases where the record of data is limited, or observed rates are low, the stochastic process involved in sampling from modeled distributions may not adequately reflect the uncertainty around the estimated input parameters. We present a conceptually simple, but computationally demanding, method for generating variance in incidence through the use of bootstrapping; for each subsample, a regression model is fitted, and the simulation model is run repeatedly sampling from the fitted model. Stochasticity is introduced at two levels; data for fitting the regression, and sampling from the fitted model. We illustrate this hybrid approach using Norwegian stroke records to generate stroke incidences with age, sex, and location, in a simulation model made to analyze travel time, queuing, and time to treatment in regional stroke units.

The packing process is a critical post-harvesting activity in table grape industry. Workers in packing stations are hired under seasonal contracts because of product seasonality and operations labor intensity. Seasonal workers, however, are usually characterized by inconsistent performance, high turnover and experience variation which leads to low productivity and high waste. Few mathematical models were used for evaluating fresh products packing operations, but in a deterministic nature which hinders the complexity and dynamics of the business processes. Hence, a hybrid Discrete Event Simulation (DES) and Agent-Based Modelling (ABM) are employed to evaluate a set of seasonal recruiting policies in a large grape packing station. The paper aims to investigate the impact of workers experience on packing operations efficiency. The model outcomes demonstrate the improvement in operations efficiency and total running cost (about 20% savings) that can be achieved when applying optimal recruiting policies that reduce labors variations.

Personal data underpins much of our digital lives with recommenders proposing products and services that themselves result from personal usage data analysis. The ownership and use of personal data is central to much current legislation in an ever changing commercial environment. Design fictions are utilized here to explore a future where consumers (termed ‘prosumers’) take ownership of their personal data and offer it to a marketplace of data buyers. Personal data trading offers a disruptive opportunity to empower the end user to realize the value of their own data using new electronic trading platforms that aggregate data in response to buyer requirements. An personal data exchange-based trading platform is described where data content classification determines a selling facade (or ‘persona’). Exploratory agent-based and system dynamics models emerge and are used to examine behavior, market dynamics and process flow within this fictional trading scenario.

We present the design of a hybrid control system for an Unmanned Aerial Vehicle (UAV) used for data collection from wireless sensors. We postulate a restrictive scenario where a low-cost processor is in charge of both flying the UAV and resolving data communication. This raises the need for safe trade-off of computing resources between stability and throughput, adapting to unpredictable environment changes. We present a strategy where a supervisory controller implements an adaptive relaxation of the sampling period of the UAV regulation controller to favor communication tasks. To guarantee stability under period switching we update the discrete-time control law with suitable gains. The resulting system comprises continuous, discrete-time and discrete-event dynamics, including event-based adaptation of the discrete-time controller. We show how the DEVS modeling and simulation framework can support a full simulation-based design, verification and validation process, featuring a seamless composition of the underlying hybrid domains.

Simulation and simulation optimization have primarily been used for static system design problems based on long-run average performance measures. Control or policy-based optimization has been a weakness, because it requires a way to predict future behavior based on current state and time information. This work is a first step in that direction with a focus on congestion measures for queueing systems. The idea is to fit predictive models to dynamic sample paths of the system state from a detailed simulation. We propose a two-step method to dynamically predict the probability of the system state belongs to a certain subset and test the performance of this method on two examples.

Simulations are often used for the design of complex systems as they allow to explore the design space without the need to build several prototypes. Over the years, the simulation accuracy, as well as the associated computational cost has increased significantly, limiting the overall number of simulations during the design process. Therefore, metamodeling aims to approximate the simulation response with a cheap-to-evaluate mathematical approximation, learned from a limited set of simulator evaluations. Kernel-based methods using stationary kernels are nowadays wildly used. However, using stationary kernels for non-stationary responses can be inappropriate and result in poor models when combined with sequential design. We present the application of a novel kernel-based technique, known as Deep Gaussian Processes, which is better able to cope with these difficulties. We evaluate the method for non-stationary regression on a series of real-world problems, showing that it outperforms the standard Gaussian Processes with stationary kernels.

We investigate maximum likelihood (ML) and ordinary least squares (OLS) methods to fit a continuous piecewise linear (PL) intensity function for non-homogeneous Poisson processes. The estimation procedures are formulated as convex optimization problems that are highly tractable. We also study the model mis-specification issues for ML and OLS in settings where the point process is non-Poisson or the underlying intensity is not piecewise linear. Through the computational study, the performances of ML and OLS estimators are exhibited. The customer arrival process to a large U.S. bank call center is studied using our methods.

In this paper we present the stochastic co-kriging methodology (SCK) for approximating a steady-state mean response surface based on outputs from both long and short simulation replications performed at selected design points. We provide details on how to construct an SCK metamodel, perform parameter estimation, and make prediction via SCK. We demonstrate numerically that SCK holds the promise of providing more accurate prediction results at no additional computational effort by only externally adjusting the simulation runlength and number of independent replications of simulations through the experimental design of the simulation study.

In many situations, e.g., simulation optimization and input uncertainty quantification, we need to assess the system performance at a large number of alternative inputs. Since each simulation run could be computationally expensive, statistical emulator could efficiently use the simulation budget to estimate the system performance. This paper proposes a new emulator for stochastic simulation, called asymmetric kriging (AK), which can be used to emulate the distribution of stochastic simulation outputs at each input point. Different from existing methods in the simulation literature, our approach does not require strong assumptions on either the functional form of the response surface or the normal distribution of the simulation estimation error. Numerical studies indicate the efficacy of our approach compared to alternative methods in the literature.

Cellular regulatory mechanisms are typically governed by biochemical reaction networks. Discrete stochastic models are widely used in computational systems biology to analyze such networks. Often, the models involve a large number of highly uncertain parameters and many interacting chemical species. However, one is often interested in observing the output of one, or a few of the species rather than the entire network. Simulating the complete reaction network is inefficient in such cases. This paper explores the use of surrogate models to learn partial stochastic biochemical reaction networks and enable fast near-instant evaluation. The efficacy of the proposed method is demonstrated on a model from the systems biology literature.

Simulation is used often in the preliminary design phase when there are fewer control parameters thus making it feasible to use tangible visual analysis. Tangible user interfaces and interactions are realized by deploying real objects representing control parameters. Those objects can be moved and rotated in order to interact with computer and direct the visual analysis. We can take advantage of new technologies, such as Microsoft HoloLens, to provide a mixed reality based system for tangible visual analysis. Instead of using generic tokens, as the current state of the art does, we use semantic representatives that can function without augmentation. We also introduce the iconic view, integrated within a coordinated multiple views tool, which depicts input parameters. The iconic view can be used as an alternative to the tangible interface for input parameter specification. The preliminary results indicate that manipulating simulation parameters in a less abstract way helps the experts.

We propose a stochastic mesh approach to portfolio risk measurement under the nested setting in which revaluation of the portfolio value requires simulations. While stochastic mesh was originally proposed as a tool for American option pricing, we are interested in estimating via simulation the risk of the portfolio in our context. We establish the asymptotic properties of the stochastic mesh estimator for portfolio risk. In particular, we show that the estimator is asymptotically unbiased and consistent, and its mean squared error (MSE) converges to zero in a rate of gamma^(-1), where gamma is the effort required to simulate the sample paths. This rate of convergence is the same as that under the non-nested setting. The proposed method allows for path dependence of financial instruments in the portfolio. Preliminary numerical experiments show that the proposed method works reasonably well.

Stochastic simulation is an indispensable tool in operations and management applications. However, simulation models are only approximations to reality, and typically bear discrepancies with the generating processes of real output data. We investigate a framework to statistically learn these discrepancies under the presence of data on past implemented system configurations, which allows us to improve prediction using simulation models. We focus on the case of general continuous output data that generalizes previous work. Our approach utilizes (a combination of) regression analysis and optimization formulations constrained on suitable summary statistics. We demonstrate our approach with a numerical example.

Morris's elementary effects method (MM) has been known as a model-free factor screening approach especially well-suited when the number of factors is large or when the computer model is computationally expensive to run. In this paper, we propose the controlled Morris method (CMM) that acts in a sequential manner to keep the computational effort down to a minimum. The sequential probability ratio test-based multiple testing procedure adopted by CMM enables to identify the factors with significant main and/or interaction effects while controlling Type I and Type II familywise error rates at desired levels. A numerical example is provided to demonstrate the efficacy and efficiency of CMM.

Stratified sampling and Latin hypercube sampling (LHS) reduce variance, relative to naive Monte Carlo sampling, by partitioning the support of a random vector into strata. When creating these estimators, we must determine: (i) the number of strata; and, (ii) the partition that defines the strata. In this paper, we address the second point by formulating a nonlinear optimization model that designs the strata to yield a minimum-variance stratified sampling estimator. Under a discrete set of candidate boundary points, the optimization model can be solved via dynamic programming. We extend this technique to LHS, using an approximation of estimator variance to obtain strata for the domain of a multivariate function. Empirical results show significant variance reduction compared to using equal-probability strata for LHS or naive Monte Carlo sampling.

The mean time to failure (MTTF) of a stochastic system is often estimated by simulation. One natural estimator, which we call the direct estimator, simply averages independent and identically distributed copies of simulated times to failure. When the system is regenerative, an alternative approach is based on a ratio representation of the MTTF. The purpose of this paper is to compare the two estimators. We first analyze them in the setting of crude simulation (i.e., no importance sampling), showing that they are actually asymptotically identical in a rare-event context. The two crude estimators are inefficient in different but closely related ways: the direct estimator requires a large computational time because times to failure often include many transitions, whereas the ratio estimator entails estimating a rare-event probability. We then discuss the two approaches when employing importance sampling; for highly reliable Markovian systems, we show that using a ratio estimator is advised.

Virtual performance is a class of time-dependent performance measures conditional on a particular event occurring at time t0 for a (possibly) nonstationary stochastic process; virtual waiting time of a customer arriving to a queue at time t0 is one example. Virtual statistics are estimators of the virtual performance. In this paper, we go beyond the mean to propose estimators for the variance, and for the derivative of the mean with respect to time, of virtual performance, examining both their small-sample and asymptotic properties. We also provide a modified K-fold cross validation method for tuning the parameter k for the difference-based variance estimator, and evaluate the performance of both variance and derivative estimators via controlled studies. The variance and derivative provide useful information that is not apparent in the mean of virtual performance.

Fuzzy Cognitive Mapping (FCM) represents the `mental model' of individuals as a causal network equipped with an inference engine. As individuals may disagree or evidence be insufficient, causal links may be assigned a range rather than one value. When all links have range, the massive search space is a challenge to running simulations. In this paper, we presented, implemented, and evaluated a new approach to identify which ranges are important and simplify models accordingly. Our approach uses a factorial design of experiments, implemented using parallelism to offset its high computational cost. Our implementation (including our new Python library for FCM) is freely available on a third-party repository. Our evaluation on three previously published models shows that our approach can simplify almost half of a model under common settings, and runs within seconds on entry-level hardware for small FCMs. Further research is needed on simplifying the few FCMs having many links.

We study the problem of the Monte Carlo estimation of the right tail of the distribution of the sum of correlated log-normal random variables. While a number of theoretically efficient estimators have been proposed for this setting, using a few numerical examples we illustrate that these published proposals may not always be useful in practical simulations. In other words, we show that the established theoretical efficiency of these estimators does not necessarily convert into Monte Carlo estimators with low variance. As a remedy to this defect, we propose a new estimator for this setting. We demonstrate that, not only is our novel estimator theoretically efficient, but, more importantly, its practical performance is significantly better than that of its competitors.

Developing metamodels for quantiles can be inaccurate when the input estimates of the quantiles used to fit the model are noisy. In this paper, a multiple response model is developed to jointly model the quantile with a correlated and less-noisy expectation to improve the fit and predictions from the quantile metamodel. We first extend the standard stochastic Gaussian process model to the multi-response case and then use a simple m-design-point example to analytically study the benefits of the joint model over the single model. Several other numerical experiments are also conducted, and the results show that the joint model can provide better performance and thus improve quantile predictions.

Simulation from the tail of the multivariate normal density has numerous applications in statistics and operations research. Unfortunately, there is no simple formula for the cumulative distribution function of the multivariate normal law, and simulation from its tail can frequently only be approximate. In this article we present an asymptotically efficient Monte Carlo estimator for the tail of the multivariate normal distribution. The estimator leverages upon known asymptotic approximations. In addition, we generalize the notion of asymptotic efficiency of Monte Carlo estimators of rare-event probabilities to the sampling properties of Markov chain Monte Carlo algorithms. Regarding these new notions, we propose a simple and practical Markov chain sampler for the normal tail that is asymptotically optimal. We then give a numerical example from finance that illustrates the benefits of an asymptotically efficient Markov chain Monte Carlo sampler.

A simplified stochastic simulation model is often used to guide real-time decision making for a complex real system, such as scheduling decisions for semiconductor production. To provide reliable guidance, we propose a simulation calibration framework. We first develop a spatial-temporal metamodel to estimate the system dynamic behaviors at different settings of calibration parameters. Then, assisted by the metamodel, we introduce a calibration model so that the dynamic behaviors of the calibrated simulation model match with those of the real system. Thus, for any feasible decisions, the calibrated simulation model can predict the future outputs for the real system and deliver prediction intervals.

Advancements in computer technology have resulted in improved computing capability and software functionality. Concurrently, in the simulation community demand to study complex, integrated systems has grown. As a result, it is difficult to perform model exploration or optimization simply due to time and resource limitations. Metamodeling offers an approach to overcome this issue; however, limited study has been made to compare the methods most appropriate for simulation modeling. This paper presents a contemporary comparison of methods useful for creating a metamodel of a simulation model. For comparison we explore the performance of a complex system dynamics model of a community hospital. In our view several characteristics of hospital operations present an interesting challenge to explore and compare the well-known competing methods. We consider three dimensions in our comparison: fit quality, fitting time, and results interpretability. The paper discusses the better performing methods corresponding to these dimensions and considers tradeoffs.

In this paper we propose a novel misspecification test for simulation metamodels. It is a consistent test that helps to assess the adequacy of simulation metamodels. The test statistic we construct is shown to be asymptotically normally distributed under the null hypothesis that the metamodel is correct, while diverging to infinity at a rate of sqrt(n), where n is the test sample size if the given metamodel is inadequate. Furthermore, as a by-product, we construct confidence intervals for mean squared errors of the metamodels. Preliminary numerical studies show that the test works quite well and has good finite-sample properties.

We develop a new unbiased estimation method for Lipschitz continuous functions of multi-dimensional stochastic differential equations with Lipschitz continuous coefficients. This method provides a finite variance estimator based on a probabilistic representation which is similar to the recent representations obtained through the parametrix method and recursive application of the automatic differentiation formula. Our approach relies on appropriate change of variables to carefully handle the singular integrands appearing in the iterated integrals of the probabilistic representation. It results in a scheme with randomized intermediate times where the number of intermediate times has a Pareto distribution.

A simplified simulation model is often used to guide the decision-making for a real complex stochastic system. To faithfully assess the mean performance of the real system, it is necessary to calibrate the simulation model. Existing calibration approaches are typically built on the summary statistics of simulation outputs and ignore the dynamic information carried by the detailed sample paths. In this paper, we develop a new calibration approach incorporating the detailed output sample paths in a sequential manner. Our theoretical development and empirical study demonstrate that we can efficiently use the simulation resources and achieve better calibration accuracy by exploring the system dynamic behaviors.

Bias due to input modelling is almost always assumed negligible and ignored. It is known that increasing the amount of real-world data available for modelling input processes causes this form of bias to decrease faster than the variance due to input uncertainty. However, this does not mean bias is irrelevant when considering the error in a simulation performance measure caused by input modelling. In this paper we present a response surface approach to bias estimation in simulation models along with a diagnostic test for identifying, with controlled power, bias due to input modelling of a size that would be concerning to a practitioner.

Quantiles are often employed to measure risk. We combine two variance-reduction techniques, conditional Monte Carlo and Latin hypercube sampling, to estimate a quantile. Compared to either method by itself, the combination can produce a quantile estimator with substantially smaller variance. In addition to devising a point estimator for the quantile when applying the combined approaches, we also describe how to construct confidence intervals for the quantile. Numerical results demonstrate the effectiveness of the methods.

In this paper, we introduce a sequential procedure for the Multi-Objective Ranking and Selection (MOR&S) problems that identifies the Pareto front with a guaranteed probability of correct selection (PCS). In particular, the proposed procedure is fully sequential using the test statistics built upon the generalized sequential probability ratio test (GSPRT). The main features of the new proposed procedure are: 1) a unified framework, the new procedure treats the multi-objective problems in the same way as the single-objective problems; 2) an indifference-zone-free formulation, the new procedure eliminates the necessity of indifference-zone parameter; 3) asymptotically optimality, the GSPRT achieves asymptotically the shortest expected sample size among all sequential tests; 4) general distribution, the procedure uses the empirical likelihood for generally distributed observation. A numerical evaluation demonstrates the efficiency of the new procedure.

Principal-agent problems study contracts for goods or services that a principal (seller) should offer an agent (buyer). The goal is for the principal to optimize the quantity and price in the contract offered to an agent with uncertain demand, where the principal has estimated a distribution for the agent's demand. The agent's demand distribution can be discrete or continuous. A deterministic optimization solution to the discrete distribution problem delivers a contract with price and quantity options targeted towards each possible demand realization. When the demand distribution is continuous, the optimal contract becomes a continuous function of the demand space. This paper introduces a sample average approximation to the continuous distribution problem using methods for solving the discrete distribution problem. We explore using numerical results an example motivated by carbon capture and storage systems.

Optimizing the process parameters with respect to the future environmental conditions is an immediate challenge for environment-sensitive process manufacturing industry to achieve more consistent production quality. In this paper, we propose a simulation-based quality variance control system consisted of three core components: an indoor environment calibration module, a quality prediction module and a simulation engine. We then demonstrate the use of this system by analyzing a typical manufacturing process consisted of four sub-processes. The studies show that the proposed system can achieve better performances by integrating future indoor environment calibration module than that of without such module. In addition, the simulation-based method can provide more acceptable outcome which outperforms the collaborative filtering algorithm. Such system is feasible to be applied in real industry scenarios which are sensitive to environmental changes to precisely control the quality variances.

When it comes to building a successful public bike share such as the NYC CitiBike system, there are many important questions to answer. There is no shortage of work being done on finding the most efficient way to redistribute bikes to stations. Equally important to how to distribute bikes is the question of when to redistribute bikes. Redistribute too infrequently and customers become frustrated, resulting in decreased revenue. Redistribute too frequently and the cost of redistribution becomes prohibitively high. In this piece of research we attempt to find the optimal time to call for the redistribution of bikes to minimize cost and retain maximum membership.

We focus on simulation optimization algorithms that are designed to accommodate noisy black-box functions on mixed integer/continuous domains. There are several approaches used to account for noise which include aggregating multiple function replications from sample points and a newer method of aggregating single replications within a "shrinking ball." We examine a range of algorithms, including, simulated annealing, interacting particle, covariance-matrix adaption evolutionary strategy, and particle swarm optimization to compare the effectiveness in generating optimal solutions using averaged function replications versus a shrinking ball approximation. We explore problems in mixed integer/continuous domains. Six test functions are examined with 10 and 20 dimensions, with integer restrictions enforced on 0%, 50%, and 100% of the dimensions, and with noise ranging from 10% to 20% of function output. This study demonstrates the relative effectiveness of using the shrinking ball approach, demonstrating that its use typically enhances solver performance for the tested optimization methods.

Simulated systems are often described with a variety of models of different complexity. Making use of these models, algorithms can use low complexity, "low-fidelity" models or meta-models to guide sampling for purposes of optimization, improving the probability of generating good solutions with a small number of observations. We propose an optimization algorithm that guides the search for solutions on a high-fidelity model through the approximation of a level set from a low-fidelity model. Using the Probabilistic Branch and Bound algorithm to approximate a level set for the low-fidelity model, we are able to efficiently locate solutions inside of a target quantile and therefore reduce the number of high-fidelity evaluations needed in searches. The paper provides an algorithm and analysis showing the increased probability of sampling high-quality solutions within a low-fidelity level set. We include numerical examples that demonstrate the effectiveness of the multi-fidelity level set approximation method to locate solutions.

We study a random search algorithm for solving deterministic optimization problems in a black-box scenario. The algorithm has a model-based nature and finds improved solutions by sampling from a distribution model over the feasible region that gradually concentrates its probability mass around high quality solutions. In contrast to many existing algorithms in the class, which are population-based, our approach combines random search with a two-time-scale stochastic approximation idea to address a certain ratio bias inherent in these algorithms and uses only a single candidate solution per iteration. We prove global convergence of the algorithm and carry out numerical experiments to illustrate its performance.

There has been recent interest, and significant success, in adapting and extending ideas from statistical learning via Gaussian process (GP) regression to optimization via simulation (OvS) problems. At the heart of all such methods is a GP representing knowledge about the objective function whose conditional distribution is updated as more of the feasible region is explored. Calculating the conditional distribution requires inverting a large, dense covariance matrix, and this is the primary bottleneck for applying GP learning to large-scale OvS problems. If the GP is a Gaussian Markov Random Field (GMRF), then the precision matrix (inverse of the covariance matrix) can be constructed to be sparse. In this paper we show how to exploit this sparse-matrix structure to extend the reach of OvS based on GMRF learning for discrete-decision-variable problems.

Optimization of complex real-time control systems often requires efficient response to any system changes over time. By combining pattern search optimization with a fast estimated Gaussian Process model, we are able to perform global optimization more efficiently for response surfaces with multiple local minima or even dramatic changes over the design space. Our approach extends the pattern search for global optimization problems by incorporating the global and local information provided by an additive global and local Gaussian Process model. We further develop a global search method to identify multiple promising local regions for parallel implementation of local pattern search. We demonstrate our methods on a standard test problem.

The field of simulation optimization has seen algorithms proposed for local optimization, drawing upon different locally convergent search methods. Similarly, there are numerous global optimization algorithms with differing strategies to achieve convergence. In this paper, we look specifically into meta-model based algorithms that stochastically drive global search through an optimal sampling criteria evaluated over a constructed meta-model of the predicted response considering the uncertainty of the response. We propose Trust Region Based Optimization with Adaptive Restart (TBOAR), a family of algorithms that dynamically restarts a trust region based search method via an optimal sampling criteria derived upon a meta-model based global search approach. Additionally, we propose a new sampling criteria to reconcile undesirable adaptive restart trajectories. This paper presents preliminary results showing the advantage of the proposed approach over the benchmark Efficient Global Optimization algorithm, focusing on a deterministic black box simulator with a d-dimensional input and a one-dimensional response.

In this study, we consider the robust Ranking and Selection problems with input uncertainty. Instead of adopting the minimax analysis in the classical robust optimization, we propose a novel method to approach this problem from the perspective of multi-objective optimization and Pareto optimality. More specifically, the performances of each design under various scenarios are reformulated as multiple objectives, and in this case, robust Ranking and Selection becomes a multi-objective Ranking and Selection. In order to determine the number of simulation replications to various scenarios of each design, a bi-level convex optimization is formulated by maximizing the surrogate of the large deviation rate function of the probability of false selection. Numerical results show the efficiency of the proposed procedure (PR-OCBA) compared with other methods.

In a Ranking and Selection problem, the objective of allocation efficiency is vital in deriving the rule. However, most of these objectives do not have a close form. Due to the high cost of a direct approximation, several cheap but biased substitutes were applied to simplify the problem. This simplification however could potentially affect the optimality of original problem and therefore influence its finite performance. Fortunately, due to the increasing accessibility of parallel hardware (e.g. GPU), a direct approximation is becoming more tractable. Thus, we want to test the performance of an allocation rule based on an unbiased and direct approximation, expecting an acceleration on the performance. In this paper, we target on one of the popular objective, the Possibility of Correct Selection (PCS). Numerical experiments were done, showing a considerable improvement in finite performance of our algorithm comparing to a traditional one.

We consider a new ranking and selection problem in which the performance of each alternative depends on some observable random covariates. The best alternative is thus not constant but depends on the values of the covariates. Assuming a linear model that relates the mean performance of an alternative and the covariates, we design selection procedures producing policies that represent the best alternative as a function in the covariates. We prove that the selection procedures can provide certain statistical guarantee, which is defined via a nontrivial generalization of the concept of probability of correct selection that is widely used in the conventional ranking and selection setting.

We consider a ranking and selection problem where sampling of two alternatives at once is required for learning about the true performances of the individual alternatives. The true performance of an alternative is defined as its average probability of outperforming the other alternatives. We derive and numerically compare four different solution approaches. Two Knowledge Gradient sampling policies are compared with a pure exploration policy and with a knockout tournament. The knockout tournament serves as a natural benchmarking approach with respect to pairwise comparisons, and determines the sampling budget provided to the other approaches. Our numerical results show that the Knowledge Gradient policies outperform both knockout tournament and pure exploration, and that they lead to significant improvements already at a very small number of pairwise comparisons. In particular we find that a nonstationary Knowledge Gradient policy is the best of the considered approaches for ranking and selection with pairwise comparisons.

This paper considers the problems of identifying the best of a discrete set of alternatives for each of a set of correlated problem instances that can be described by a set of continuous features, and where the performance of a particular alternative on a particular problem instance can only be estimated from noisy samples. A possible application is in manufacturing, where we would like to identify the best dispatching rule to be used depending on shop floor conditions, and performance is estimated via stochastic simulation. We propose three myopic sequential sampling methods to collect information, and in particular focus on an efficient estimation of the expected improvement which requires integration over the continuous space of problem instances. Empirical tests show that our method of estimating expected improvement is indeed more efficient than standard Monte Carlo integration, and that our method significantly outperforms a recently published alternative sampling method.

Expected improvement (EI) is a leading algorithmic approach to simulation-based optimization. However, it was recently proved that, in the context of ranking and selection, some of the most well-known EI-type methods cause the probability of incorrect selection to converge at suboptimal rates. We investigate a more recent variant of EI (known as "complete EI") that was proposed by Salemi, Nelson, and Staum (2014), and summarize results showing that, with some minor modifications, complete EI can be made to achieve the optimal convergence rate in ranking and selection with independent Gaussian noise. This is the strongest theoretical guarantee available for any EI-type method.

This study formulates and solves the design problem for a master-worker architecture dedicated to the implementation of a parallelized simulation optimization algorithm. Such a formulation does not assume any specific characteristic of the optimization problem being solved, but the way the algorithm is parallelized. In particular, we refer to the master-worker paradigm, where the master makes sampling decisions while the workers receive solutions to evaluate. We identify two metrics to be optimized: the throughput of the workers in terms of the number of evaluations per time unit, and the lack of synchronization between the master and the workers. We identify several design parameters: number of workers (n), the buffer size for each worker and for the master and the sample size m as the number of solutions used by the master for sampling decisions at each iteration. Numerical experiments show optimal designs over randomly generated simulation optimization algorithm instances.

Epidemiological models have tremendous potential to forecast disease burden and quantify the impact of interventions. Detailed models are increasingly popular, however these models tend to be stochastic and very costly to evaluate. Fortunately, readily available high-performance cloud computing now means that these models can be evaluated many times in parallel. Here, we briefly describe PSPO, an extension to Spall's second-order stochastic optimization algorithm, Simultaneous Perturbation Stochastic Approximation (SPSA), that takes full advantage of parallel computing environments. The main focus of this work is on the use of PSPO to maximize the pseudo-likelihood of a stochastic epidemiological model to data from a 1861 measles outbreak in Hagelloch, Germany. Results indicate that PSPO far outperforms gradient ascent and SPSA on this challenging likelihood maximization problem.

We empirically evaluate the finite-time performance of several simulation-optimization algorithms on a testbed of problems with the goal of motivating further development of algorithms with strong finite-time performance. We investigate if the observed performance of the algorithms can be explained by properties of the problems, e.g., the number of decision variables, the topology of the objective function, or the magnitude of the simulation error.

Comparing with the well-studied unconstrained ranking and selecting problems in simulation, literatures on constrained ranking and selection problems are relatively fewer. In this paper, we consider the problem of ranking the top-m designs subjected to stochastic constraints, where the design performance of the main objective as well as the constraint measures can only be estimated from simulation. Using the optimal computing budget allocation framework, we derive an asymptotically optimal allocation rule. The effectiveness of the suggested rule is demonstrated via numerical experiments.

Ranking and Selection (R\&S) procedures are designed for selecting the best among a finite set of systems using stochastic simulation, guaranteeing the quality of the final selection. Instead of assuming a known lower bound on the difference between the best and others, we consider the probably approximately correct (PAC) selection formulation, which ensures a high quality solution with high probability for all configurations. In this paper, we present a new fully sequential selection procedure, called the Envelope Procedure (EP), which accommodates a variety of sampling rules that, together with a carefully defined termination condition, ensures a PAC selection. A particular sampling rule that achieves good efficiency is proposed. We compare the efficiency of the EP with some existing procedures in numerical experiments, and the results show that the EP saves considerable computational effort in many problem configurations.

Many classical sequential procedures model the partial sum difference process between two competing alternatives as a Brownian motion process. In this paper, the marginal probability of eliminating the best alternative is considered while modeling the partial sum difference process. We adaptively allocate the total amount of the probability of incorrect selection to every time point where the comparisons between alternatives are conducted and set the continuation regions to ensure the marginal probability of eliminating the best alternative does not exceed the probability assigned to each time point t. We show by examples that under our framework, the procedure can be easily developed for both indifference-zone (IZ) and IZ free formulations.

A widely acknowledged challenge in ranking and selection is how to allocate the simulation budget such that the probability of correction selection (PCS) is maximized. However, there is yet another challenge: when the input distributions are estimated using finite real-world data, simulation output is subject to input uncertainty and we may fail to identify the best system even using infinite simulation budget. We propose a new formulation that captures the tradeoff between collecting input data and running simulations. To solve the formulation, we develop an algorithm for two-stage allocation of finite budget. We use numerical experiment to demonstrate the performance of our algorithm.

Simulation-based Optimization (SbO) assumes that the simulation model is valid, and that the probability distributions used therein are accurate. However, in practice, the input probability distributions (input models) are estimated by sampling data from the real system. The errors in such estimates can have a profound impact on the optimal solution obtained by SbO. The existing two-stage framework for SbO under computational budget constraint considers only the stochastic uncertainty. In our variant, we consider the input model parameter uncertainty as well. Our algorithmic procedure is based on the stochastic kriging metamodel-assisted bootstrapping with an efficient global optimization technique which sequentially searches the optimum and incorporates Optimal Computational Budget Allocation (OCBA). This framework is also used for determining tighter worst case bounds of the SbO with input uncertainty. The proposed framework is illustrated with the M/M/1 queuing model.

We consider simulation optimization in the presence of input uncertainty. In particular, we assume that the input distribution can be described by some continuous parameters, and that we have some prior knowledge defining the probability distribution for these parameters. We then seek the simulation design that has the best expected performance over the possible parameters of the input distributions. Assuming correlation of performance between solutions and also between input distributions, we propose modifications of two well-known simulation optimization algorithms, Efficient Global Optimization and Knowledge Gradient with Continuous Parameters, so that they work efficiently under input uncertainty.

Online change detection has many applications, ranging from finance and manufacturing, to security and computer vision. Designing a change detector for use in a given domain can be very time consuming, and model-based algorithms often require knowledge of the underlying stochastic model. To address these issues, in this work we explore a supervised learning approach to a change detector. We implement a gradient based procedure to find the optimal parameters for a change detector. We demonstrate the methodology on both synthetic and real world data for classifying 3D laser range image data in real-time.

Motivated by big data applications, we consider unconstrained stochastic optimization problems. Stochastic quasi-Newton methods have proved successful in addressing such problems. However, in both convex and non-convex regimes, most existing convergence theory requires the gradient mapping of the objective function to be Lipschitz continuous, a requirement that might not hold. To address this gap, we consider problems with not necessarily Lipschitzian gradients. Employing a local smoothing technique, we develop a smoothing stochastic quasi-Newton (S-SQN) method. Our main contributions are three-fold: (i) under suitable assumptions, we show that the sequence generated by the S-SQN scheme converges to the unique optimal solution of the smoothed problem almost surely; (ii) we derive an error bound in terms of the smoothed objective function values; and (iii) to quantify the solution quality, we derive a bound that relates the iterate generated by the S-SQN method to the optimal solution of the original problem.

Consider the context of integer-ordered bi-objective simulation optimization, in which the feasible region is a subset of the integer lattice. We propose a retrospective approximation (RA) framework to identify a local Pareto set that involves solving a sequence of sample-path bi-objective optimization problems at increasing sample sizes. We apply the epsilon-constraint method to each sample-path bi-objective optimization problem, thus solving a sequence of constrained single-objective problems in each RA iteration. We solve this deterministic optimization problem using the SPLINE algorithm, thus exploiting gradient-based information. In early RA iterations, when sample sizes are small and standard errors are relatively large, we provide only a rough characterization of the Pareto set by making the number of epsilon-constraint problems a function of the standard error. As the RA algorithm progresses, the granularity of the characterization increases until we solve only as many epsilon-constraint problems as there are points in the local Pareto set.

We study a direct Monte-Carlo-based approach for computing the worst-case expectation of two multidimensional random variables given a specification of their marginal distributions. This problem is motivated by several applications in risk quantification and statistics. We show that if one of the random variables takes finitely many values, a direct Monte Carlo approach allows to evaluate such worst case expectation with $O(n^{-1/2})$ convergence rate as the number of Monte Carlo samples, $n$, increases to infinity.

Conditional Value-at-Risk (CVaR) is a widely used metric of risk in portfolio analysis, interpreted as the expected loss when the loss is larger than a threshold defined by a quantile. This work is motivated by situations where the CVaR is given, and the objective is to find the portfolio with the largest or smallest quantile that meets the CVaR constraint. We define our problem within the classic stochastic multi-armed bandit (MAB) framework, and present two algorithms. One that can be used to find the portfolio with largest or smallest loss threshold that satisfies the CVaR constraint with high probability, and another that determines the portfolio with largest or smallest probability of exceeding a loss threshold implied by a CVaR constraint, also at some desired probability level.

We provide a unified framework to treat the asymptotic analysis for the non-batched quantile sensitivity estimators of Fu et al. (2009), Liu and Hong (2009), and Lei et al. (2017). With only mild differences in regularity conditions and proofs, asymptotic results including strong consistency and a central limit theorem are established for all three estimators. Simulation results substantiate the theoretical analysis.

Constructing a building is a very complex process including a vast number and variety of participants, tasks, materials and elements. Interdependencies between the on- and off-site logistics frequently lead to an error-prone and chaotic process. However, good communication can reduce the effects by enabling the different actors to react. Simulation is a classic approach to understand complex problems and has been widely applied in the construction field. Nevertheless, the existing construction simulation approaches still do not consider the entire construction logistics process. This paper proposes a concept to model and simulate on- and off-site construction logistics, to facilitate the understanding of the boundary-spanning dependencies of both on- and off-site domains. A framework is provided, including simulation fundamentals (i.e. logistic BIM model), simulation data preparation (i.e. process pattern definition) and implementation of a scenario-based simulation. Finally, a discussion of the possible benefits and improvements of the proposed approaches is provided.

In construction, major capital projects are in need of a visualization method for scheduling and integrating the spatial dimensions with the time dimension. Traditional scheduling methods are limited to the time dimension, and can be used to visualize the critical path of schedules and to compare the criticality of activities. However, they do not consider the spatial constraints. This paper describes a method for developing 4D simulation to visualize the criticality of project activities considering the requirements of the levels of detail. The 4D visualization interface shows the criticality of activities linked to components with color coding based on the total float of each activity. Important benefits can be achieved in supporting decision-making associated with understanding the spatio-temporal constraints related to multiple contracts. Furthermore, the proposed method is useful for filtering, viewing critical and near critical activities, and comparing schedules. The method is tested in a hydroelectric powerhouse case study.

Discrete event simulation (DES) has been used in academia to solve complex construction management problems for decades. However, it has not gained widespread adoption within the industry. A considerable effort to increase the acceptance of DES in construction has been carried out. However, it has been heavily focused on implementing DES models, neglecting both a detailed conceptualization of the real system and an understanding of stakeholders’ requirements. Conceptual modeling (CM) is a fundamental step in DES which helps build stakeholders’ trust in the resulting model. We argue, then, that it is vital to develop a CM framework for the construction domain to improve the acceptance of DES in the industry. We propose a domain-specific CM framework for construction application. This framework will be formulated to suit the complex environment of construction projects.

This study aims to assess the carbon dioxide emission of carwash construction operation by applying computer-based simulation modeling. Discrete event simulation (DES) servs as an effective way to functionally evaluate carbon dioxide emission of a typical carwash construction. The DES model makes it possible to efficiently simulate and analyze carbon dioxide emission factors. This in turn requires project participants to find ways to decrease inverse environmental effects of the projects to respond to the increasing concern on such issues. This study is carried out by studying the process, data analysis, and simulation of the construction operation using Visio, Excel, and EZStrobe computer software. This simulation tool is practical for project participants and will offer them an idea of carbon dioxide emission in more realistic scenarios during construction operation.

The total energy involved in building construction and operation can be divided in two parts, Operational Energy and Embodied energy. Many researches have focused on simulating operation energy but the embodied energy is not much discussed for simulation or alternative selection purposes. A lot of complex variables need to be quantified and analyzed for optimization of embodied energy. Building Information Model (BIM) is used as a platform to hold the data as a context aware, intelligent and interactive database This paper discusses the means for realizing the embodied energy data in the BIM model through linking the model with external databases thus laying a foundation for the data exchange and retention needed to perform the simulations in the next phase of research. Plug-ins are developed to calculate the embodied energy for different scenarios. Finally, a case study is conducted of a simplified manufacturing plant to implement the proposed methodology.

The operation of ready-mix concrete (RMC) plants is strongly affected by the complex stochastic nature of the concrete production and delivery processes. The delivery compliance in the different building sites is one of the most significant aspects for both customer satisfaction and resource allocation in the RMC industry. To improve the decision-making process in the RMC industry, a multi-method (using a hybrid Agent-based and Discrete-Event simulation modeling technique) simulation model was created for a multinational company that produces and markets cement and ready-mixed concrete using software Arena, Rockwell Automation. The main contribution of this work is the novel approach to designing travel time to the construction sites, improving its validity and allowing us to optimize the production and delivery program. Results obtained in this project were used in the decision making process regarding the effective use of capital in installed capacity while minimizing risk.

Construction sites comprise constantly moving heterogeneous resources that operate in close proximity of each other. The sporadic nature of field tasks creates an accident prone physical space surrounding workers. Despite efforts to improve site safety using location-aware proximity sensing techniques, major scientific gaps still remain in reliably forecasting impending hazardous scenarios before they occur. In the research presented in this paper, spatiotemporal data of workers and site hazards is fused with a quantifiable model of an individual’s attitude toward risk to generate proximity-based safety alerts in real time. In particular, a worker’s risk index is formulated and coupled with robust hidden Markov model (HMM)-based trajectory prediction to approximate his/her future position, and detect imminent contact collisions. The designed methodology is explained and assessed using several experiments emulating interactions between site workers and hazards. Preliminary results demonstrate the effectiveness of the designed methods in robustly predicting potential collision events.

The construction industry records the highest rate of accidents amongst all industrial sectors. Accidents occur due to shortcomings in the identification of unsafe conditions before an activity is executed. Building Information Modeling (BIM) is a promising development in the construction industry, which can be applied to considerations relating to occupational health and safety issues. Only a few research activities have earlier focused on the use of building information models to identify safety hazards. The present work deals with the integration of the processes of risk assessment in a model-based environment. Based on algorithms that have been developed, a systematic identification and categorization of hazards - based on a building model - has been devised. The hazards are specifically identified with regard to object, process, and environment, and documented in the context of an object-oriented database. This paper will show the process of input data acquisition for generating the knowledge base.

The objective of this study was to investigate the underlying mechanisms of vulnerability in complex construction projects using simulation experiments. Specifically, two hypotheses related to project vulnerability were tested: (1) project schedule performance is negatively correlated with vulnerability; (2) the level of project vulnerability is positively correlated with project exposure to uncertainty and organizational complexity. In the proposed dynamic network simulation methodology, construction projects are modeled as heterogeneous meta-networks. Project vulnerability is assessed by the decrease in meta-network efficiency due to uncertainty-induced perturbations. Project schedule deviation is used as a measure for quantifying the impacts of vulnerability on project performance outcomes. The proposed simulation methodology was implemented in three case studies of real-world construction projects. Monte-Carlo simulation experiments were conducted under different simulation scenarios consisting of varying levels of uncertainty and project planning strategies to test the hypotheses.

Construction industry has been constantly lagging behind in terms of efficiency and productivity growth. Simulation modeling can be used to improve the productivity of construction workflow processes through modeling uncertainties and stochastic events that may negatively impact project cost and schedule. In the research presented in this paper, mobile sensors coupled with machine learning techniques are used for ubiquitous data collection and human activity recognition (HAR), which will constitute the key input parameters of process simulation modeling. To assess the designed methodology, an experiment is carried out which replicates a warehouse quality control operation. Smartphones mounted on human bodies are used to collect multi-modal time-motion data. Support vector machine (SVM) is then applied to classify workers’ and inspectors’ activities, and activity durations are subsequently extracted. Finally, a simulation model is built using the output of the HAR phase and rigorously validated and used to analyze workflow processes, productivity, and bottlenecks.

Construction simulation is used to analyze uncertainties inherent to project activities and variations in work packages. However, existing simulation systems often fail to meaningfully contribute to the decision-making process due to their inability to evolve with changing project conditions. Equipping simulation models with sensing and reality capture technologies has been investigated as possible remedies to this problem. This, however, requires meticulous effort to procure, set up, operate, synchronize, and calibrate peripheral devices for data collection, transmission, and mining. Furthermore, sensor readings are often noisy and imperfect. The chaos theory explains how small variations in sensor readings used as simulation model input can lead to relatively large volatility in the output even in simple linear systems. This paper investigates a scientific methodology for generating more stable simulation models using an evolutionary algorithm that produces clean datasets by processing and significantly reducing noise in imperfect data obtained from consumer-grade sensors.

Recent breakthroughs in BIM and ADC technologies promise innovative solutions to bridge the information gaps between the digital models and real construction site. These solutions promote the collaboration between digital, spatial and physical construction. Cyber-physical systems offer a tight integration between real physical and virtual “cyber” models. This collaborative approach supports the digital transformation in construction domain. A cyber-physical framework is proposed to provide consistent relationships and allow bidirectional data flow. In the framework the recognition of objects successes by linking physical objects to the digital product models using RFID. Next, these objects are equipped with global positions data and pinned to semantic and functional enrichment construction places. The results are objects at a level of “smartness” with enhanced digital capabilities and the ability of context-awareness. The cyber-physical objects are embedded in the process models in order to support tracking activities and facilitate process monitoring and control close to real-time.

Building Management Systems (BMS) monitor and control smart buildings. We are witnessing a trend of BMS becoming increasingly sophisticated and delivering more advanced services. The downside of this is that the complexity of BMS is also increasing, thus, making smart buildings vulnerable to various malfunctions and faults. We anticipate that reliability of smart buildings will be gaining in importance, especially for BMS of critical buildings, such as hospitals or buildings that host emergency services or store sensitive materials or technologies, whose unreliable operation could have catastrophic consequences. The assistance, however, is in the large amounts of easily available data that implies possibilities for development of highly accurate reliability models. Cloud computing can be also utilized to support collaborative sharing and benefitting from each other building’s data. To utilize all of the above stated, we have developed a cloud-based BMS reliability analysis framework that we describe and illustrate in this paper.

The author seeks a practical approach to complement deterministic design optimization in performance-based building design. This study investigates algorithmic processes to test and monitor the dynamic optimal control of building components. To this end, an integration of wireless data transfer equipment (nRF24L01) and a customized metaheuristic hybrid optimization algorithm (Tabu-based adaptive pattern search simulated annealing, T-APSSA) through a parametric visual programming language (VPL) interface (Rhino grasshopper®) is presented with experimental design of a responsive kinetic shading device. To demonstrate the performance of the algorithmic hybridization and early design integration, T-APSSA is compared to simulated annealing and pattern (direct) search, and two different approaches to daylight-optimized design solutions are tested: a deterministic optimization based on historical weather data and a site-specific adaptive optimization according to real-time monitoring of incident solar radiance. The suggestion of a seamless environmental building design workflow through remote data communication contributes to strengthening intelligent architectural design decisions.

To enhance the accuracy in estimating material and crew costs for steel reinforcement installation, numerous estimating tools have been developed. “Precise estimating” in general considers both lapping details and other required supporting structures while deriving the crew cost by accounting for reinforcing operations. In contrast, “rough estimating” ignores rebar lapping details in quantity takeoff and relies on industry benchmark productivity data for crew cost estimation. The distinction between “precise estimating” and “rough estimating” still lacks quantitative evidence and remains vague to both academic researchers and professional estimators. This research presents systematic comparison between the two estimating strategies with a case study of a bridge deck. A discrete event simulation tool is used to aid in estimating the crew cost in reinforcement handling and installation. The estimating results indicate that compared with the “precise estimating” approach, the “rough estimating” approach underestimates the material and crew costs by 13% and 38%, respectively.

Airports are intermodal hubs and natural interfaces between ground transport and air transport. In the current DLR project “Optimode.net”, an innovative approach is being developed to extend the management of an airport not only to airport landside and terminal processes but to go even further and incorporate feeder traffic in the management of airport processes. Thus providing travelers with a real door-to-door service and letting airport stakeholders benefit from efficient airport management. Technical core of the project is a simulation environment consisting of nine different simulation models with various simulation methods and abstraction levels. In this paper the simulation environment of a multi airport region which is used in the Optimode.net project will be described in detail and also the interaction of the different simulation modules will be explained. We will also show how this complex simulation environment is used to foster individual door-to-door travel and proactive airport management.

Aircraft boarding is a process mainly impacted by the boarding sequence, passenger behaviour and the amount of hand luggage. Whereas these aspects are already addressed in scientific research and operational improvements, the influence of infrastructural changes are only focused upon in the context of future aircraft design. The innovative Side-Slip Seat technology holds the potential for sustainably improving the boarding time by providing a wide aisle during the boarding progress. A comprehensive, validated simulation environment is used to analyse the benefits of this technology and an adapted boarding strategy is identified using evolutionary algorithms. Considering the air transportation environment (e.g. seat load), the passenger behaviour (e.g. conformance), and operational deviations (stochastic model), the Side-Slip Seat could fasten aircraft boarding by at least 20% shorter average boarding time and ensure a more stable boarding progress (smaller standard deviation).

We discuss our experience from helping an airport operator and their team of airport operations analysts to introduce discrete-event simulation alongside their existing system improvement/development toolkit. Our project looked at improving the existing baggage handling system of a major European airport run by our partner organization. This paper is divided into two separate but related discussions. First, we collate recent reflections from this project together with those from similar airport projects we have been involved in for the past decade. We discuss a few observations that we believe should help to convince airport operators to develop more in-house skills around both analytical and simulation modeling and analysis, with a focus on the latter. Secondly, we describe one of the simulation components of our baggage handling project and demonstrate the effectiveness of various possibilities to improve the availability and reliability of these systems when operating under tight capacity constraints.

A flying aircraft produces two coherent rotating vortices of air in its trail. If another aircraft flies into one of these vortices, it can experience an un-commanded roll. Because of the potential risk, wake separation standards exist to significantly reduce the probability of such events. Thus, wake encounters are inherently rare. As new procedures and technologies are proposed to increase the capacity of the airspace, rare-event simulation is necessary to evaluate to the safety of proposed changes. This paper explores the performance of a rare-event splitting technique in the context of estimating probabilities of potential wake encounters. The goal is to identify good strategies for the splitting method while using a relatively simple model for the wakes. Suggestions for the choice of the level function and the locations of levels are given.

The ever-increasing demand for global air travel provokes many daunting challenges for the aviation industry in meeting future requirements. Estimation of fuel consumption plays a central role in aircraft performance evaluation, which is vital in understanding and validating key components of a future air traffic management system. This study provides a framework for a quantitative evaluation on fuel burn estimation with the objective of proposing modeling techniques to accurately assess the aircraft performance. Data from a series of continuous descent operations simulations conducted on full-flight simulator are used as reference data. Base of Aircraft Data (BADA) aircraft performance model data provided by Eurocontrol are integrated with point mass dynamics to estimate fuel flow in both clean and non-clean configurations of the aircraft. Statistical results show the model can estimate the total fuel consumption within ±6% of actual value with ±6% and ±10% for cruise and descent phases respectively.

One of the hot topics in the field of air traffic management is continuous descent operations, which is one of the efficient descent procedure to an airport. In Japan, the continuous descent operations are implemented at Kansai International Airport, but its applicability is limited to only the non-congestion period owing to the unpredictability of the trajectory and arrival time of the continuous descent operations as compared to the conventional operations. To expand the time period for the continuous descent operations, in this paper, we develop a fast-time simulator. The simulator can provide the predicted trajectory and arrival time of the continuous descent operations. The prediction accuracy is reviewed by comparison with the results of the simulation conducted on a full flight simulator. Discussions on accuracy improvement of the prediction have been provided considering available data for air traffic controllers through radar data systems.

Today’s aviation industry is faced with three conflicting goals: First, aeronautics already draws responsible for 2% of all anthropologically induced emissions (IATA 2013) and the need for reducing those emissions is no point of contention anymore. Second, the irresistible growth of air traffic demand challenges both, airport operations and air traffic flow management. And third, despite all changes, the expectations of safety are constantly increasing. To account for the need of improved, safe and climate friendly aircraft operations, we present a simulation environment, that is capable of optimizing trajectories regarding multiple criteria. Therewith we show the positive impact of 4D optimized trajectories on the air traffic flow, controller’s taskload and airspace complexity and emphasize the importance of an enhanced air traffic flow management considering dynamic sectorisation and air space configuration.

Runway expansion planning is mainly based on specific airport studies, complicating the transfer of results to another airport. Thus a normalized airport is introduced, including a set of technologies/procedures, to investigate possibilities to gain runway capacity. The advantage of this approach is the comparability of different enhancements and understanding their effectiveness under different situations. So before establishing a new runway the current runway structure can be used more efficiently, leading to optimal solutions for individual constraints. This is shown for a set of operational enhancements like runway usage strategy or sequencing, studied under different traffic mixes as well as arrival-departure ratios. Hereby the main focus of this paper is set to dependent parallel runway configurations.

The Federal Aviation Administration is funding and encouraging new concepts for improving air traffic flow management (TFM) decision-making. The resulting automation capabilities will need to be operationally integrated into the existing air traffic management system. Understanding how a new capability will interact with existing system components is challenging because of the range of possible real-world situations. Although there are fast-time traffic simulation tools available for modeling the impact of TFM actions, they are often developed as stand-alone tools, which are not extensible to work with other external capabilities for conducting integration studies or quantifying benefits. To address this gap, we have built a fast-time, distributed simulation platform integrating state-of-the-art traffic simulator and allows the plug-in of advanced TFM prototypes so their interactions can be studied. This paper discusses the requirements and the necessary components for building this platform and then use a TFM integration case study to demonstrate its utility.

This paper deals with the improvement of the robustness of heuristic solutions for aviation systems affected by uncertainty when the resolution of conflicts is implemented. A framework that includes the use of optimization and simulation is described which in turn generates pseudo-optimal schedules. The initial solution is progressively improved by iteratively evaluating the uncertainty in the generated solutions and calibrating in accordance with the objective function. Simulation is used for testing the feasibility of a solution generated by an optimization algorithm in an environment characterized by uncertainty. The results show that the methodology is able to improve solutions for the scenarios with uncertainty, thus making them excellent candidates for being implemented in real environments.

We present the concept and initial results of using fast-time air traffic simulation modeling to assess requirements for convective weather forecast accuracy from Traffic Flow Management (TFM) standpoint. For strategic TFM applications, such requirements can be relaxed compared with tactical weather avoidance. A gradual increase in forecast error does not necessarily cause gradual changes in operational costs associated with a given TFM action which was implemented based on that forecast. Modest inaccuracies in forecast may not require adjustments to TFM actions; but when the discrepancy between forecast and actual weather exceeds a certain threshold, it may prompt a different TFM strategy to be initiated vs. an optimal one based on the actual weather. This, in turn, may cause a significant increase in operational costs. This paper demonstrates how such thresholds, indicative of forecast accuracy requirements for TFM, can be determined using parametric forecast accuracy changes in a series of simulations.

Quantifying the impact of an intervention or treatment in a real setting is a common and challenging problem. For example, we would like to calculate the environmental implications of aid projects in third world countries that target economic development. For causal inference problems of this kind, the Rubin causal model is one of several popular theoretical frameworks that comes with a set of algorithmic methods to quantify treatment effects. However, for a given data set, we neither know the ground truth nor can we easily increase the size of the data set. So, simulation is a natural choice to evaluate the applicability of a set of methods for a particular problem. In this paper, we report findings of a simulation study with four causal inference approaches, namely two single tree approaches (transformed outcome tree, causal tree), and two random forest versions of the former.

In mammals, touch is encoded by sensory receptors embedded in the skin. For one class of receptors in the mouse, the architecture of its Merkel cells, unmyelinated neurites, and heminodes follow particular renewal and remodeling trends over hair cycle stages from ages 4 to 10 weeks. As it is currently impossible to observe such trends across a single animal’s hair cycle, this work employs discrete event simulation to identify and evaluate policies of Merkel cell and heminode dynamics. Well matching the observed data, the results show that the baseline model replicates dynamic remodeling behaviors between stages of the hair cycle – based on particular addition and removal polices and estimated probabilities tied to constituent parts of Merkel cells, terminal branch neurites and heminodes. The analysis shows further that certain policies hold greater influence than others. This use of computation is a novel approach to understanding neuronal development.

Computer simulations are an important tool for the current research in population and evolutionary genetics. They help to understand the genetic evolution of complex processes dynamics that cannot be analytically predicted. The basic idea is to generate synthetic data sets of genetic polymorphisms under user-specified scenarios describing the evolutionary history and genetic architecture of a species. In this work, we focus on forward-in-time simulations which represent the most powerful, but, at the same time, most compute intensive approach for simulating the genetic material of a population. We present a highly-optimized forward-in-time simulation library called Libgdrift, specially designed to create large sets of replicated simulations. Our simulation library uses code optimizations such as spatial locality and a two-phase data compression approach which allow fast simulation executions, while reducing memory storage. Results show that our proposal can improve the performance reported by well-known simulation software.

Industrial symbiosis provides several positive aspects regarding energy and material efficiency, but it is also a challenging concept due to additional dependencies in a production cluster. As a consequence the risks inherent to these partnerships are high and thus the concept is not widely used. The project ESProNet supports the analysis and assessment of industrial symbiosis in a given cluster via simulation and research of altered scenarios. This paper contains the first steps in the project including the ontology to breakdown the complex problem as well as the modeling approach for the components in a new library based on material and energy balances. A proof of concept with the simulation of an industrial symbiosis cluster containing a server providing waste heat to an office building shows a great potential with up to 20 % less energy input for heat.

As share of renewable sources in the energy sector is increasing, the energy production and distribution network’s centralized structure is changing to numerous small-scale distributed networks. Energy consumers in the residential sector are increasingly becoming energy producers by adopting rooftop photo voltaic (PV) systems. However, increasing rooftop PV adoption has contributed to diminishing revenues for utility companies. This paper describes an agent-based model that has been developed to help utility companies better understand the impacts of consumers’ preferences and behaviors on adding renewable sources to their energy mix. Experimental results demonstrate that including both consumers and utility companies as stakeholders can help the utilities alleviate revenue losses due to increasing rooftop PV adoption while meeting their renewable energy expansion targets.

There is a national need for a quick and easy building energy performance assessment system of existing buildings, without resorting to dynamic building energy simulation tools which usually require significant cost, time and expertise. In this study, the authors report the development of a building energy profiling system which is based on Artificial Neural Network (ANN) models. The ANN models were made by a series of EnergyPlus pre-simulations sampled by a Monte Carlo technique. The MBE and CVRMSE between EnergyPlus and ANN models are 1.53% and 7.82%, respectively. It is concluded that the profiling system requires minimalistic inputs and provides accurate energy performance assessment of a given building.

The present article examines the reasons for a steady growth in the number of registered floods and demonstrates the necessity to develop strategies ensuring integrated land and water resources management. A modern methodology of developing rules for the use of reservoirs is considered, and the basic criteria are formulated, which form the basis for planning the flow out of the reservoir. The effectiveness of the hydrotechnical structures’ cascade compensating management and control mode is established. The objective of developing an hybrid simulator based on the present model was set and reached. Computer simulation boundary conditions and mathematical apparatus are formulated. The developed software modules flowcharts are presented, which allows to control the water level in a reservoir and at all the subsequent river reaches depending on the predetermined hydrographs of water inflow and water pass through hydrotechnical structures. Furthermore, the given software operating principle is demonstrated.

The modeling of changes in landscape cover is a task that helps to predict the development of the territory, which may be useful, for example, in planning infrastructure development (Singh et al. 2015). In practice it is also useful to consider other development scenarios of the study area. In this paper we consider the task of handling the simulation by alternative scenarios using cellular automata and Markov chains. Approach to the preparation of initial data for modeling by an alternative scenario is proposed, the conditions for the correctness of the modification are considered, and a modification algorithm is proposed that allows maintaining correctness when partial modification conditions are met. The approach is tested on the problem of land cover changes forecasting of Portugal territory in the framework of the project LANDYN.

To determine the budget needed for the staff needed for mental health services required by new government regulations, a simulation model of the process was built, verified validated, and used to identify where mandated delivery times were not being met, where staff should be reallocated. In addition to the obvious benefits of this approach, a less obvious benefit was that upon the further examination that occurred as part of the discovery process needed to build the model, additional opportunities were found for providing better care with less resources. A final benefit of this work was the potential recognized by the Chief Financial Officer that the approach utilized here could be used throughout the network to rationalize staffing levels, and thus make it possible to provide more, better or timelier outcomes with the same resources throughout the network.

In recent years, all acute hospitals in the UK have experienced unprecedented emergency department waiting times and hospital bed pressures. The consequences are overcrowded emergency departments, ambulance shortages, cancelled elective operations, low staff morale and financial penalties. To deal with the increasing numbers of patient admissions and delayed discharges hospitals must turn now to modelling and simulation to help increase their flexibility and ability to deal with demand variation. Hospitals face several issues that reduce their flexibility including the need for extreme value-for-money and specialization of care. This talk presents three ED case studies undertaken by an analytics team in the UK. The paper considers the impact of the work and challenges arising from their experiences of simulation modelling in acute hospitals. Final thoughts consider the future of ED simulation.

Walk-in clinics have grown in popularity in the United States as a substitute for traditional medical care delivered in primary care clinics and emergency rooms. Walk-in clinics offer an affordable option for basic medical services when compared to a hospital emergency room or an urgent care clinic. This type of medical facility simplifies the health care process for many patients with non-life threatening conditions since no previous appointments are required to see a provider. However, the open access nature and lack of patient scheduling can lead to long wait times for patients or long periods of idle time for providers. In this paper, we derive a discrete event simulation model to study pure walk-in clinics where patients are served without appointments. A case study is discussed that considers a walk-in clinic located in central Texas. The computational study provides useful insights that are applicable to any walk-in health care facility.

Renal failure concerns progressive loss of kidney function. Renal Replacement therapy (RRT) is a costly, long-running process that includes several decision points in different stages. Small changes in the protocol can impact significantly the expenditures and healthcare outcomes. Unfortunately, policy makers have very little support for benchmarking improvement alternatives. The existing models are designed to fit certain applications with preset parameters and design choices which do not match with the requirements of a policy analysis. A generic approach is required to analyze the effects of different design options adjustable to finer scales. To remedy this, this paper describes a novel toolkit for evaluating renal replacement policies, containing a parametrized colored Petri-Net which can be configured for the specifics of local settings. The model is made available for open access to overcome the non-replicability issue of existing models.

Managing the supply and demand of blood for transfusions is a complicated problem that many hospitals encounter since the blood supply is dependent on donations while the demand is not. Moreover, recent research also shows that transfusing older blood into patients may lead to increased mortality. This raises the issue of whether transfusing fresher blood can be achieved without jeopardizing blood availability. In this paper, we build a simulation model to study the impact of racial and ethnic diversity in combination with a blood allocation threshold. We show that for many diverse cities like New York City that the increased diversity of the population can lead to a large mismatch between blood supply and demand.

In this paper is presented the design of an active ankle prosthesis with a single degree of freedom (DoF) actuated by a linear actuator. The analysis is focused in the plantar base that shall allow the impacts absorption in gait cycle with the aim to replace the human lower limb of a person who has suffered ankle amputation. For the design it has been simulated the natural motion of a human ankle which trajectory is known for previous studies based in a biomechanical analysis of the same. For the static structural analysis has been used finite elements software ANSYS, obtaining data of stress, deformation and security factor which allow choose and couple properly the ideal linear actuator for the kind of patient and, the properly thickness of the plantar base with which the prosthesis will be built to avoid mechanic failures.

The spread of diseases such as the Middle East Respiratory Syndrome (MERS) and avian influenza inflicts a significant socioeconomic problem, and highlights the need for a systematic analysis of disease spread patterns. In Korea, however, most domestic research utilize equation-based approaches that treat the entire country as a single entity and thus have limited applicability. In this study, we propose an agent-based disease spread model that reflects not only the nature of the disease, but also the structural and statistical characteristics of each sub-region’s population. Based on the 2010 Census data, we develop a population model of Suwon city in Korea. The spread of MERS disease and various response strategies were analyzed based on the contact network based on the socioeconomic activities of residents. The proposed model is expected to play an important role in formulation of effective disease-related policies, reflecting the mobility and socio-economic structure of today’s urban society.

The U.S. is in the grips of a devastating opioid and heroin co-epidemic affecting nearly all socio-economic populations at great human (~7,800 new users/day) and financial ($78.5 billion/year) costs but with no obvious solution. We describe recent work and challenges to develop, integrate, and use several analytic multi-scale simulation models of these epidemics to develop insight into the epidemic’s complex underlying dynamics, generate causal hypotheses, and inform effective policy interventions. We developed preliminary agent-based, differential equation, network spread, and cellular automata models that reasonably replicate at multiple scales the past 17 years of this epidemic’s growth and spread at town, county, state, and national levels. Results suggest that some current approaches are unlikely to be very effective, some in fact may worsen the epidemic, and ultimately only certain combinations and sequences of policies are likely to have value, with important implications on both model architecture and policy optimization.

This study aims to use agent-based simulation as a tool to illustrate the importance of human behavior in the dynamics of vector-borne disease spread. For this, a baseline compartmental model was developed and, based on it, four different scenarios considering human behavior were proposed: two assuming the whole population adopts the same behavior and two assuming each individual has his/her own behavior. Paired t-test was used to compare the proposed scenarios with the baseline, based on two output responses from the simulation experiments: total number of infected people and duration of the epidemic. Results from the data analysis indicate that behavior is an important factor and, as such, it must be further investigated and included in infectious-disease spread models to obtain more accurate results. As a final remark, we presented possible explanations to why human behavior has been neglected in many epidemiological models up to now.

This paper presents a framework aimed at making decisions on sample size and power that optimize expected net benefits of clinical trials that incorporate health and cost variables. Two-stages standard modeling and simulation was developed. The hybrid framework was populated with prior information on benefits of a clinical trial that compare magnetic resonance imaging vs. computerized axial tomography in a trial of acute ischemic stroke diagnostic, combined with an assumption on willingness to pay per health gain, correlated distributions, exclusivity and per patient-time cost. As results, an estimated optimal expected net benefits of €3.8M at optimal power of 34% and sample size of 1800 were calculated. Emphasis on main and sensitivity analysis results are also presented.

Simulation has been used for modeling in healthcare for many decades. Ranging from the modeling of physiological processes to group dynamics to the modeling of strategic and system-wide models of healthcare provision, simulation promises to be an effective approach to analyze healthcare operations. Effective application of simulations in healthcare operations requires that simulation deal with wide variability and unpredictability in workflow processes, the complexity of healthcare organizations and enables the participation of human experts in the modeling and operations processes. In this paper, based on requirements drawn from a participatory simulation with healthcare practitioners, we define a hybrid, composable approach to healthcare simulations. Both the participatory simulation and the composable simulation are applied in the context of the New Karolinska Solna hospital in Sweden, a highly specialized new hospital. Results point to the need to accounting for variability in workflow processes and integration with existing IT infrastructure in hospitals.

Utilization evaluation for healthcare facilities such as hospitals and nursing homes is crucial for providing high quality healthcare services in various communities. In this paper, a data-driven simulation framework integrating statistical modeling and agent-based simulation (ABS) is proposed to evaluate the utilization of various healthcare facilities. A Bayesian modeling approach is proposed to model the relationship between heterogeneous individuals’ characteristics and time to readmission in the hospital and nursing home. An ABS model is developed to model the dynamically changing health conditions of individuals and readmission/discharge events. The individuals are modeled as agents in the ABS model, and their time to readmission and length of stay are driven by the developed Bayesian individualized models. An application based on Florida’s Medicare and Medicaid claims data demonstrates that the proposed framework can effectively evaluate the healthcare facility utilization under various scenarios.

This study describes a simulation model that was used to evaluate a proposed workload score. The score was designed to assist in triaging patients into the hospital services of the Division of Hospital Internal Medicine at Mayo Clinic in an effort to more equitably balance workload among the division’s provider teams (or services). The first part of this study was the development of a score, using Delphi surveys, conjoint analysis, and optimization methods, that accurately represents provider workload. A simulation model was then built to test the score using historical patient data. Preliminary simulation results reported the proportion of time that each provider team spent working at or above “maximum utilization,” as defined by Mayo Clinic experts. The model yielded a 12.1% decrease (on average) in the proportion of time provider teams spent at or above maximum utilization, while simultaneously displaying a more balanced workload across provider teams.

Successful merging or consolidation of interdependent healthcare clinics have been shown to have benefits with regards to decreasing operation costs while maintaining patients’ quality of care. In order to achieve a successful merger or integration of clinics, an analysis of the effects of integrating patient flows should occur. This is especially important when the merger of clinics involves a transition into a new facility. We utilize a discrete event simulation model to study the effects of integrating three interdependent musculoskeletal clinics, Orthopedics, Rheumatology, and Radiology, into a new facility in advance of implementation. Through use of the simulation, unexpected bottlenecks in the check-in process are identified and the effects of implementing new patient flows, supported by Real-Time Location System (RTLS) technology, are analyzed.

Caregivers of patients with chronic diseases are undergoing a daily burnout in their lives. Although respite care seems a promising solution, no quantitative analysis has yet been provided to demonstrate its positive impact. In this article, we propose (i) a new model of caregivers' burnout evolution based on Markov chain and machine learning to model health state evolution, and (ii) a multi-agent based simulation approach to describe the burnout evolution of caregivers and impact of respite structures on the system. Optimal capacity of respite structures is obtained through a design of experiment. Several management strategies are also tested (collaboration between structures, reservation of beds for emergent cases). Key performance indicators considered are quality of service and costs. Results show a positive impact of respite services on both quality of service and costs. The model also show a trade-off between quality of service and costs when bed reservation policies are used.

The saturation of the Emergency Department services is mostly due to admission of non-urgent or minor-urgency patients who represent a high percentage of admitted patients in the service. We propose a model for scheduling the entry of these non-critical patients into the Emergency Department which may be helpful for the management of the service dealing with the current growing demand for emergency medical care. We hypothesize that a relocation of these non-critical patients in the expected input pattern, initially provided by actual historical data from the hospital, can lead to an improvement in waiting times for all patients, and therefore, to an improvement in the quality of service from the point of view of the service users, as it could avoid long waiting times in the service. Simulation is used to show and evaluate the effect of applying the proposed scheduling model.

This paper focuses on optimizing patient scheduling at a breast cancer center for two types of patients: follow up and consult patients. Follow up and consult patients have different service times and follow different care pathways. The objective of this paper is to sequence the patients such that minimum average flow time is achieved for each patient type. A simheuristic framework is developed by integrating MATLAB, Simio, and Excel. Unlike existing simulation-optimization (SO) approaches that target the simulation model controls, this framework tries to optimize a data table in the simulation environment. This simulation evaluation (SE) approach could iteratively input patients arrival table into the simulation model, and obtains the expected performance of the system as a reference to generate the next solution. The obtained results from this framework are further analyzed by comparing five heuristic appointment scheduling methods.

Calibration is a vital part of the modeling and simulation process and denotes the determination of parameter values by estimating them from comparison between simulation results with reference data. During the last decades a lot of algorithms have been developed for this purpose which are able to fit mentioned parameter values generically without information about the modeled system or the simulation. Especially for stochastic simulation models these routines very often require thousands of iterative simulation executions which, in case of large agent-based models, might be too time intensive. In this paper we illustrate a real world example for such a problem and present a solution for it based on probability theory. Hereby we not only calibrate the desired parameters, but also find a measure for the quality of the fit as well. By presenting this example we want to motivate modelers to analyze agent-based models to save costly calibration time.

The use of dashboards to aid hospital decision makers in managerial and clinical decisions is well documented in the literature, though few broach the challenging subject of combining cost measurement with user satisfaction and building layout optimization. This paper presents an innovative dashboard in a 3D environment, providing decision makers with simulation capabilities using agent based simulation, allowing examination of their facility and the impact of policy, process and layout changes on patients and finances. An example is presented for an Emergency Department, wherein the presented dashboard revealed that the costs of constructing additional triage rooms would produce no benefit to patients; rather, a change in the process would be more beneficial compared with the existing situation. It is concluded that the developed dashboard allows users to make comparisons between multiple scenarios and visualize data in an intuitive format, allowing for decision makers to optimize their facility and operations.

Hospital planners want to build the right capacity based on expected demand. Overbuilding means overspending and building beds that become underutilized. Underbuilding means insufficient beds, causing longer bed queues. Placing inpatients in semi-private beds reduces capital investment cost by building fewer rooms, or potentially fewer floors, but this option usually reduces patient satisfaction compared to placing inpatients in all private beds. Additionally, studies show inpatients with private bed experience have shorter lengths of stay. This paper describes how simulation modeling can help reduce capital costs by determining an appropriate number of semi-private rooms that ensure sufficient capacity yet care for most inpatients with a private room experience. Results from a recent inpatient simulation project indicates possible reduction of private rooms by 25 percent yet only requiring inpatient placement in semi-private beds 5 percent of time. Further analysis shows this result is only possible when inpatient capacity is precisely determined.

A simulation model was created to model the traffic flow in the operating room. A key research challenge in operating room design is to create the most efficient layout that supports staff and patient requirements on the day of surgery. The simulation allows comparison of base model designs to future designs using several performance measures. To develop the model, we videotaped multiple surgeries in a set of operating rooms and then coded all activities by location, agent and purpose. Our current analysis compares layouts based on total distance walked by agents, as well as the number of contacts, measured as the number of times agents must change their path to accommodate some other agent or physical constraint in the room. We demonstrate the value and capability of the model by improving traffic flow in the operating room as a result of rotating the bed orientation.

This study conducts an operational analysis of horizontal evacuation in a hospital, using a crowd density model to simulate patient evacuation. The proposed discrete event simulation model considers three types of patients assigned with specific moving speed and areas from which to evacuate; the moving speed also varied according to the area’s crowd density. We evaluated the effect of the ratio and transfer priority for each patient type on evacuation time in the ward. During night shifts, the evacuation time increased with the ratio of type 1 patients transferred by two nurses with sheets or blankets. The results showed only slight differences in evacuation times when transfer priority changed between type 1 and type 2 patients. The hospital required only one additional staff member to shorten the evacuation time during night shifts

Health care system is one of the most critical units in case of disasters. Floods cause an increase of emergency patient flow that may overwhelm hospital resources. In this paper, we present a simulation model that evaluates health care emergency plan and assesses the resilience of the Ile-de-France region in case of a major flood. We combined in the model the health care process with a Markov chain flood model. The results can be used to elaborate an optimized strategy for evacuation and transfer operations. We provide a case study on three specialties and quantify the impact of several flood scenarios on the health care system.

Many Canadian hospitals run at or near capacity, frequently experiencing congestion due to surges in demand. “Surge Protocols” that formally define when and what kind of operational steps can be taken to alleviate congestion are routinely in use. Decisions across the hospital, regarding bed capacity and allocation, staffing levels, and the surgical block schedule influence the frequency and severity of congestion, which in turn manifests in high bed occupancy, delayed admissions, crowded Emergency Department, surgical cancellations and increased use of surge protocols. A generic, data-driven, discrete event simulation is developed to help hospitals assess the impact of hospital wide decisions and surge policies on congestion. The model has been developed in cooperation with two hospitals, has been validated at a third and is continuing to be applied at additional hospitals.

Virtual factory models can help improve manufacturing decision making when augmented with data analytics applications. Virtual factory models provide the capability of simulating real factories and generating realistic data streams at the desired level of resolution. Deeper insights can be gained and underlying relationships quantified by channeling the simulation output data to an external analytics tool. This paper describes integration of a virtual factory prototype with a neural network analytics application. The combined capability is used to create a neural network capable of predicting the expected cycle times for a small job shop. The capability can adapt by retraining the neural network whenever the production circumstances change significantly. The trained neural network can be used for functions such as order promising and can support factory management. The analytical and adaptive combination represented by the virtual factory integrated with the neural network thus supports the move towards smart manufacturing.

High-dimensional Markov decision processes are often solved using online sampling-based methods such as Monte Carlo tree search. Problems that involve rare catastrophic events with large negative rewards can be very challenging for such approaches. The difficulty arises because the common policies for expanding a search tree are not able to handle rare large negative rewards appropriately, leading to high variance estimates of value. This paper studies the impact of reward structure with rare catastrophic events on the performance of common algorithms and proposes enhanced tree policy strategies for Monte Carlo tree search. Numerical experiments suggest that our proposed methods can significantly improve performance on a variety of problem domains ranging from stochastic multi-armed bandits to aircraft rerouting problems.

In the robot swarm, each robot can freely form swarm with others to share information. Although particle swarm optimization (PSO) has been demonstrated to outperform Q-learning and evolutionary algorithms, less study is conducted to characterize various swarm intelligence (SI) algorithms and evaluate their performances for robot swarm learning. In this research, we select three representative SI algorithms include bat algorithm (BA), PSO, grey wolf optimizer (GWO) according to their learning strategies. These three algorithms are implemented in a distributed manner and compared under various number of robots (NR) and communication ranges (CR). The simulation results demonstrate that: 1) PSO outperforms BA and BA outperforms GWO in general, 2) GWO performs better than PSO and BA under large NR and long CR, 3) increasing NR and CR can significantly improve the performance of GWO. These results can shed lights on selection of optimal SI algorithms for robot swarm under various scenarios.

Traffic congestion leads to waste of time and has tremendous impact on the environment. To reduce traffic congestion, vehicle to vehicle (V2V) and vehicle to infrastructure (V2I) communication have been explored and implemented. This paper focuses on integrating features of gap acceptance behavior proposed in V2V and road segment occupancy & intelligent traffic light phasing of V2I. Three separate simulation models have been presented in this study. A design of experiment generated thirty scenarios to capture the overall traffic flow performance in terms of total time and waiting time in the system, WIP and system efficiency. Results revealed that the proposed V2I simulation model in which routing was based on traffic light availability decreased the waiting time, total time and WIP significantly compared to baseline and smart traffic light vehicle flow model. The routing principle in the latter two models was based on space availability approach.

Mining a discrete event simulation model from data has always been a big challenge, which is related to the problem of system inference in systems theory. D2FD (Data to Fuzzy-DEVS model) method can be used to discover a discrete event simulation model. This method not only provides a way of data mining but also integrates process mining with the modularity, frequency, timing aspects and event data. This paper presents a mature tool applying D2FD method. This tool is implemented as an available and dedicated plug-in within the open-source process mining toolkit ProM. The simulation tool SimStudio is embedded in this plug-in and it can simulate Fuzzy-DEVS atomic and coupled model. Two case studies of real life processes, taken from Rabobank Group ICT and Dutch Employee Insurance Agency, are analyzed to evaluate this tool.

Automated vehicles have been under heavy development in major auto and tech companies and are expected to release into market in the foreseeable future. However, the road safety of these vehicles remains a concern. One approach to evaluate their safety is via on-track experimentation, but this requires gigantic costs and time investments. This paper discusses a sequential learning approach based on kriging models to reduce the experimental runs and economize on-track experimentation. The approach relies on a heuristic simulation-based gradient descent procedure to search for the best next test scenario. We demonstrate our approach with some numerical test cases.

Port capacity is determined by three major infrastructural resources namely, berths, yards and gates. The advertised capacity is constrained by the least of the capacities of the three resources. While a lot of attention has been paid to optimizing berth and yard capacities, not much attention has been given to analyzing the gate capacity. The gates are a key node between the land-side and sea-side operations in an ocean-to-cities value chain. The gate system under consideration, located at an important port in an Asian city, is a multi-class parallel queuing system with non-homogeneous Poisson arrivals. It is hard to obtain a closed form analytical approach for such a system. In this paper, we describe an application of simulation techniques in analyzing the performance of gate operations. Further, we develop an optimization model that is integrated with simulation techniques to suggest efficient lane management policies for an outbound gate system.

The various processes of truck loading at a fuel filling plant are interrelated which makes it complex to analyze and improve the performance of the filling plants. This paper describes these processes and presents a software tool developed in FlexSim for modelling the drivers' activities, analyzing different decision scenarios, and optimizing the filling processes at fuel terminals. The simulation model has been used to identify bottlenecks and evaluating opportunities for performance improvement. This paper reports the application of the model to a lubricant filling plant for analyzing and supporting decision making on the assignment of trucks to bays, and on the availability of the lubricants on the fuel arms at the lading bays.

This paper describes a detailed stochastic simulation model integrated with a transactional database to model operations in an empty container depot. Empty container depots are found ubiquitously in supply chains around the world but there has been virtually no quantitative research done to assess operational policies nor layout designs. In this work we determine the performance of operational policies related to the stacking and retrieval of empty containers to derive recommendations for policy improvements and, in future work, the yard layout design. A simulation model was chosen as the proper tool to address these aims because of the uncertain nature and complex handling actions of an empty container depot. Results thus far show that policies concerning remarshaling and retrieval strongly influence the efficiency of the depot operations in terms of the truck turnaround times of trucks, as well as the utilization of the resources which include yard cranes and personnel.

The life cycle of a container terminal includes four important life stages: design, implementation, operation and optimization. In order to accomplish one of these stages it is crucial to use the appropriate approaches and tools. Two essential ingredients that help to accomplish the life stages of container terminal are simulation and emulation. In this paper the reader is guided through the maturity process of the container terminal, presenting the simulation and emulation approaches and tools applied to support each life stage.

Designing the storage yard in a container terminal is a major step towards efficient terminal operations. The problem may cover both long-term decisions, such as selecting the yard layout and the material handling equipment, and short-term decisions, such as assigning storage space and dispatching and routing the material handling equipment. We present a discrete-event simulation model to investigate the best yard layout in terms of block position, number and capacity in a straddle carrier-based transshipment hub bearing a perpendicular yard layout. The simulation model is highly detailed and it captures the blocking, locking and further waiting conditions occurring during real-time operations. Numerical experiments carried out for a real container terminal show how the model may easily support the operations manager in choosing the block design that minimizes the waiting times of straddle carriers and the makespan of the integrated container discharge/loading process.

There are many designing factors affecting the traffic efficiency in an automated container terminal; dynamically, the traffic efficiency is also influenced by the nature of job sequence in the specific container terminal, the number of vehicles deployed, and respective yard planning strategies. Therefore, it is difficult to analyze such complexity from an analytical queuing model; however, challenges arise in the simulation study such as how to effectively model the impact of the critical designing factors as well as the decision rules. In this study, we model the traffic system as a network of servers that represent both paths and junctions, for which the service rates are dynamically adjusted according to the respective states and decision rules. The model is implemented with O2DES.Net, an open-structured and modularized modeling framework. Numerical experiments illustrate the effectiveness of developed models, with an application of the AGV network for an automated container terminal.

In picker-to-parts order picking the traveling of the order picker between the storage locations account for approximately 50 % of the overall time. Reducing travel times can therefore substantially improve the productivity. Hereby current research has almost exclusively focused on minimizing the travel distance and assumed a constant velocity of the order picker. However transported item weight can significantly influence the velocity and consequently travel times as well. Hence the paper at hand analyzes to which extent travel times vary in dependence of item weights in the warehouse. New weight class based storage assignment policies are investigated. Their aim is to locate the items in the storage area according to their weight so that the heaviest items are collected at the end of the order picker’s tours. Agent-based simulation experiments confirm that the new policies can significantly reduce travel times.

Simulation models are proposed for modeling Shuttle Vehicle-type Mini-load storage and retrieval systems (SVM-AS/RSs). The SVM-AS/RS is a fast shuttle vehicle-type mini-load automated storage and retrieval system designed to provide for storage and sorting functions including inventory buffer before shipping, picking, assorting, palletizing, or merging. The systems considered in this study consist of lightweight shuttle vehicles installed on each level, storing and retrieving lifters, layer conveyors connecting lifters and shuttle vehicles, and incoming and outgoing aisle conveyors. To analyze its performance, simulation models are performed taking the relationships of the storage locations and load sequences into consideration. It is emphasized that AS/RS operations of the mini-load system can be confirmed by simulation experiments. In addition, the key performance indicators (KPIs) from the simulation analysis can be used to understand and validate its efficiency and effectiveness under different layouts.

We consider in this paper multi-deep automated storage/retrieval systems, where the cell capacity is strictly greater than one load. The main advantage of this class of AS/RS is a better use of space. Its main drawback is that, in order to retrieve a desired load, it is necessary to move all the loads in front of it. This is a common characteristic of all the multi-deep automated storage/retrieval systems. The number of loads to move is given by the rank of the load to retrieve. Our objective is to provide an empirical formula of the average retrieval rank in a multi-deep automated storage/retrieval system. With this formula, it is easy to deduce the mean retrieval time. This computation is based on multiple simulations of various AS/RS models and a regression on the obtained data. The particular case of the flow-rack automated storage/retrieval system will be considered to illustrate our contribution.

The CitScale tool is a software platform for visualizing, analyzing and comparing the impacts of smart city policies based on innovative mobility concepts in urban areas. It places emphasis on new automotive vehicles aimed at reducing traffic or environmental impacts. This paper introduces this traffic simulation-based tool, and two case studies developed for different scenarios in Barcelona City are briefly presented to demonstrate the capabilities of the tool when it is combined with microscopic traffic simulation software. The first case presents an extensive evaluation of new innovative vehicles (electric vehicles, bikes and three-wheeled scooters) and mobility concepts (trip-sharing). In the second one, data provided by connected cars is analyzed in order to compare different developed navigation strategies and how they affect the city. Finally, some of the obtained results from both cases are concisely presented in order to show the potential of the proposed tool.

The team orienteering problem is a variant of the well-known vehicle routing problem in which a set of vehicle tours are constructed in such in a way that: (i) the total collected reward received from visiting a subset of customers is maximized; and (ii) the length of each vehicle tour is restricted by a pre-specified limit. While most existing works refer to the deterministic version of the problem and focus on maximizing total reward, some degree of uncertainty (e.g., in customers' service times or in travel times) should be expected in real-life applications. Accordingly, this paper proposes a simheuristic algorithm for solving the stochastic team orienteering problem, where goals other than maximizing the expected reward need to be considered. A series of numerical experiments contribute to illustrate the potential of our approach, which integrates Monte Carlo simulation inside a metaheuristic framework.

The strong global competition in the shipbuilding market forces the shipyards to focus their efforts in optimizing their system resources. Therefore, the development of efficient medium and short-term operations strategies in the shipyard block assembly process is becoming a potential tool to be more competitive. This paper introduces a heuristic simulation-based approach to address the scheduling problem for shipbuilding in a system of multi-stage production of a real world case. The main goal is to minimize the total production , assembly and waiting time of the shipbuilding process (makespan) applying different types of heuristics rules in an advanced simulation framework. The proposed simulation model allows evaluating a large amount of blocks and sub-blocks, while satisfying a large set of hard constraints. The results are generated by using data from a real shipyard. Uncertain alternative scenarios are tested and computational experiences are deeply analyzed.

The ability to quote a competitive and reliable lead time for a new order is a key competitive advantage for manufacturers and plays a significant role in customer acquisition and satisfaction. Quoting a precise and reliable lead time requires a good prediction for the flow time of a new order. This research focuses on quantifying the dependence of the flow time upon observed job shop status variables, the size of a new order, and arrival rate of future orders. An iterative fitting procedure based on stochastic kriging with qualitative factors model (SKQ), is developed to synergistically model simulation and real manufacturing data, for the prediction of a new order's flow time.

In this paper we develop a heteroscedastic t-process metamodeling approach (TP) for approximating the mean response surface implied by a stochastic simulation and performing metamodel-based optimization. We provide details on how to construct a TP metamodel, make inference and perform prediction based on TP. We show that TP can retain the attractive properties of approaches that rely on Gaussian process, but it also enjoys enhanced flexibility, at no additional computational cost. We further provide a close-form expression for the TP-based expected improvement to perform metamodel-based optimization. We compare the predictive performance of TP and stochastic kriging (SK) via an M/M/1 inspired example, and demonstrate the performance of TP and SK-based algorithms for optimizing a simple periodic review (s,S) inventory system.

We examine the problem of fitting clearing functions that estimate the expected output of a production resource as a function of its expected workload from empirical data. Unlike most regression problems, the independent variables are not directly controllable due to the presence of a planning model that controls releases to the production system, and the release decisions made by the planning model are themselves dependent on the estimated clearing function. We propose an iterative refinement procedure that uses simulation experiments to resample data from the production system as the parameters of the clearing function are iteratively updated. We compare the iterative procedure to previously used approaches with promising results.

We consider simulation-based estimation where the inputs are calibrated from predictive models generated by random forests (RFs). RF is a common technique to produce ensemble predictions by aggregating many individual decision trees. This problem arises in data-driven applications such as individualized surgery operations scheduling and supply chain management. We investigate the estimation of the output variance contributed from the noises in the input prediction, which can be used to construct output confidence intervals. In particular, we study the integration of simulation runs with a recently proposed infinitesimal jackknife estimator that can reduce the computational burden from double layers of bootstrapping. We demonstrate our scheme with an elementary numerical example.

We consider an engineer-to-order production system with unknown yield. We model the yield as a random variable which represents the percentage output obtained from one unit of production quantity. We develop a beta-regression model in which the mean value of the yield depends on the unique attributes of the engineer-to-order product. Assuming that the beta-regression parameters are unknown by the decision maker, we investigate the problem of identifying the optimal production quantity. Adopting a Bayesian approach for modeling the uncertainty in the beta-regression parameters, we use simulation to approximate the posterior distributions of these parameters. We further quantify the increase in the expected cost due to the so-called input uncertainty as a function of the size of the historical data set, the product attributes, and economic parameters. We also introduce a sampling-based algorithm that reduces the average increase in the expected cost due to input uncertainty.

The capability of imitating real-world system operations has turned simulation into an indispensable problem-solving methodology for business system design and analysis. Often, simulation supports decisions ranging from sourcing to operations to finance, starting at the strategic level and proceeding towards tactical and operational decision-making. In such a dynamic setting, the practice of simulation goes beyond being a static problem-solving exercise and requires integration with learning. This article discusses the role of learning on simulation design and analysis motivated by the needs of industrial problems. Examples are presented to describe how selected tools of machine learning can be utilized to drive learning in the design and analysis of simulations.

This paper presents an effort to assess how real-time data can be used to re-sequence jobs in a flowshop where processing times are stochastic and the objective is the minimisation of the makespan. By conducting extensive simulation experiments, we try to quantify the advantages of collecting real-time data on the actual completion times of jobs in the shop in order to re-sequence the jobs remaining to be processed. The results show that the benefit of re-sequencing is greatly influenced by the variability of the processing times on the shop floor: There is little advantage when the variability is very high but, for low and intermediate variability levels, re-sequencing provides a way to improve the performance of the initial solution. These results may serve to establish limits on the advantages of using real-time data for improving initial sequencing decisions, at least within the boundaries of our experiments.

Effective control strategies for automated guided vehicles (AGVs) are important to companies that operate flexible manufacturing systems in terms of maximizing productivity. In this paper, we design and analyze Pickup-or-Delivery-En-Route (PDER), a multiple-load AGV dispatching algorithm. PDER is a task-determination rule that enables a partially loaded vehicle traveling to a drop off destination to pickup and/or drop off loads that the vehicle would otherwise pass by en route to the original destination. We conduct a simulation-based experiment to evaluate the effectiveness of the PDER algorithm. The results indicate that PDER can produce significant positive impacts on throughput and time in system in flexible manufacturing systems utilizing multiple-load AGVs.

Snow removal operations are required to maintain roadway safety during snowy winter conditions. Reliable plans outlining the dispatching of plow trucks must be made to deliver snow removal operations on time and within budget. Historical project performance data can be used to inform and facilitate decision-making processes associated with snow removal operations. This research proposes a data-driven simulation framework for planning snow removal projects considering weather and truck-related data collected by real-time sensors. An in-house developed simulation engine, Simphony.Net, is used to simulate operations based on input information extracted from mined sensor data. This model is capable of simulating plow operations to facilitate planning at both an operational and real-time level. What-if scenarios can be generated to simulate, predict, and optimize project and resource performance. A case study conducted in Alberta, Canada is presented to illustrate the practical application of the proposed method.

After an emergency landing, it is essential to quickly evacuate the aircraft. Typically, a maximum limit of time is set, which does not depend on the number of passengers on board, the hour (day/night), or the number of inoperative emergency exits. In order to develop strategies and protocols for efficient evacuations, it is important to define all the characteristics of the aircraft, analyze multiple scenarios that may arise, conduct a comprehensive study of passengers' behavior, and consider external factors that may affect the evacuation time. Nowadays, there are flexible and powerful object-oriented simulation tools that enable the creation of realistic models, which are useful to assess evacuation strategies by studying different scenarios. In this context, this paper presents a model to analyze realistic scenarios for the evacuation of the Airbus 380, which must be done in less than 90 seconds.

In modern society, sustainable transportation practices in smart cities are becoming increasingly important for both companies and citizens. These practices constitute a global trend, which affects multiple sectors placing relevant socio-economic and environmental challenges. Moreover, uncertainty plays a crucial role in transport activities, e.g., traveling time may be affected by road works, the weather, or accidents, among others. This paper addresses a rich extension of the capacitated vehicle routing problem, which considers sustainability indicators (i.e., economic, environmental and social impacts) and stochastic traveling times. A simheuristic approach integrating Monte Carlo simulation into a multi-start metaheuristic is proposed to solve it. A computational experiment is carried out to validate our approach, and analyze the trade-off between indicators and the effect of stochasticity on the solutions.

In today’s competitive market environment, logistics has a substantial impact on companies’ performance. Improving the logistics efficiency is a main goal for many industries, especially, for those involved in car manufacturing. This work considers a real logistics problem in a car-assembly factory. The problem consists in optimizing the supply of components from the internal warehouse to the production lines, determining the best delivering routes. We describe the problem in detail, propose a mathematical programming model, and solve it with CPLEX. Afterward, to evaluate the impact of implementing the optimal routes in realistic scenarios, we apply Monte Carlo simulation and present a comparison between both solutions. The manufacturer`s Key Performance Indicators are considered for evaluating the obtained results. The study showed that the proposed solution outperformed the current one, pointing out that the optimized routes could deal with different levels of production in a more efficient and cost reductive manner.

Managers of logistics networks have the complex task of continuously maintaining their network in good operating conditions under a changing environment. Thus, they need to identify promising actions to adapt and improve the logistics network. Such actions could be the relocation of stock or the adjustment of transport relations. In order to support the managers, the authors have previously proposed a decision support system (DSS) based on discrete-event simulation (DES). The DSS automatically examines possible actions and suggests the best actions found to the managers. Since a data-driven DES approach is used for this DSS, all actions can be described as changes to a database. In this paper, an approach for modeling, integrating and executing user-generated actions into the DSS is described, in order to increase its flexibility and usability. In conclusion, the authors propose to develop a domain-specific modeling language (DSML) for modeling actions for DES models.

This paper introduces the locomotive refueling system configuration problem, which arises when railroad companies aim to improve efficiency in refueling yards through new technologies or policies. Refueling speed is important to freight railroad operational efficiency; faster refueling can increase rail network capacity without the infrastructure cost associated with new terminals or tracks. We propose a method that integrates integer programming and discrete event simulation to inform these decisions, and we demonstrate the method on data derived from industry. Specifically, the models determine the best location (denoted the “strike line”) to align trains at the refueling platform and measure the impact on refueling yard throughput associated with adopting the optimal strike lines in combination with new refueling equipment. Results using realistic parameters demonstrate a statistically significant improvement over intuitive policies.

In the context of aircraft turnaround processes, this paper illustrates how simulation can be used not only to analyze critical activities and paths, but also to generate the associated survival functions. After motivating the relevance of the topic for both airlines and airports, the paper reviews some related work and proposes the use of Monte Carlo simulation to obtain the critical paths of the turnaround process and generate the associated survival function. This analysis is performed assuming stochastic completion times for each activity in the process. A series of numerical experiments contribute to illustrate these ideas. These experiments are based on a realistic environment considering the Boeing 737-800 aircraft, although the analysis can be easily extended to any other configuration. Different levels of passengers' occupancy are analyzed, as well as two alternative designs for the turnaround stage.

In practice, suppliers fill retailers' purchase orders to the fill-rate targets to avoid the non-compliance financial penalty, or chargeback, in the presence of service level agreement. Two chargeback mechanisms -- flat-fee and linear -- have been proven to effectively coordinate the supply chain in a single-period setting. However, the mechanisms' efficiency, the incurred penalty costs necessary to coordinate the supply chain, have not been studied yet. Since retailers are often accused of treating chargeback as an additional source of revenue, this study compares the expected penalties resulted from the flat-fee or linear chargeback to shed light on the retailers' choice of mechanisms. Using experimental scenarios consisting of various demand functions, demand variabilities, and fill-rate targets, the simulation results offer counter-evidence to the accusation.

Most common sawmill log yards are operated by wheel loaders or log stackers. As the operational costs of this way of transportation are quite high, new technologies might be advantageous. This study assesses the feasibility and the requirements to a technology which allows to highly automate the log yard operations using automated storage components (ASCs). To find a reasonable size of the automated log yard for real-life applications data of a softwood sawmill’s log yard was analyzed and a simulation model was built. The results of the simulation study enable an educated assessment of the required dimension of the automated log storage for different scenarios.

Energy policies and energy prices have increasingly been influencing the demand for wood. For long-lasting profitability and sustainable growth of companies involved in the wood market, logistics for raw material supply is of crucial importance. This article addresses possible measures for supply chains in the wood-processing industry based on a five-year forecast horizon derived from a simulation study. It describes the design and implementation of a simulation model to derive strategic action recommendations for raw material supply logistics for the raw material wood. Decisions concerning the size of storage locations, the number of operators in the system and the system costs can be supported by this analysis.

Information blackouts, sudden and short-duration failure of the information flow in a supply chain, amplify the bullwhip effect in supply chains. We use a multi-echelon, discrete-event simulation, built in Arena, to observe this phenomenon. This study uses information blackouts of inventory order history for three different lengths of time, 1, 2 and 3 periods, to help supply chain managers in decision-making during and after information blackouts. Based on the increase in bullwhip effect as the result of an information blackout, managers may decide to wait out the amplification or to use a guess and replace the missing inventory order history with the last known. The latter choice employees the common manager’s heuristic of trusting the recent past to be the best predicter of the future. Our results provide supporting evidence for such managerial decisions.

Many supply chains have grown increasingly complex, which has led to the development of different facility types. One such facility is known as a post-distribution cross docking system (Post-C). In these facilities, bulk sorted product is received from various suppliers. Each product has its own destination, so the bulk package is broken, sorted by destination, and staged by destination. Typical processing includes: sort received goods by product type; break bulk and sort out goods by destination; move palletized goods to the staging areas of their respective destinations. This paper compares a global staffing policy (in which all workers may perform any task) to a dedicated staffing policy (in which groups of workers are assigned specific tasks). Through comparisons of the two models, it was found the dedicated worker model’s benefits from reduced change-over outweigh the lower worker utilization it experiences.

We analyze a two-echelon remanufacturing system which utilizes a mix of new components as well as remanufactured old components to produce a new product. We find the optimal mix of new and old components that minimizes inventory and overall cost of the system for a fixed service levels. Additionally this system is studied assuming unreliable suppliers for new components. The systems performance is analyzed using a series of dynamic equations that are developed to describe the system under study. A simulation based optimization approach is used to study various scenarios as the demand and capacity under consideration are stochastic in nature. ARENA and Opt Quest is used for updating the equations and optimizing the system respectively. Several cases are studied under the computational study to understand the impact on the system.

The integration of information systems between the various actors organizing and executing the transport of containers to seaports is slowly progressing. Transport orders are frequently characterized by high change rates causing high manual revision effort for dispatchers. Therefore, these order changes, often received shortly before the day of departure, raise the question regarding the immediate transmission of transport orders to the subsequent actors in the transport chain. This paper analyzes the impact of different order release times, which define the timing of order transmission, on order process efficiency (processing times and costs) using a multi-method simulation approach. In a case study, four actors, two focusing on transport planning and two on operative transport execution, are considered. The simulation experiments with varying order release times and change rates reveal: A late release of orders from planning to operative actors and a reduction of order changes can significantly increase order process efficiency.

Operators of parcel transshipment terminals face the challenge of sorting and transferring a large number of parcels efficiently. In order to meet customers’ expectations, short sorting intervals are required. In this paper we present a technical framework that combines metaheuristics with discrete-event simulation (DES) to provide robust solutions for problems at the operational level of parcel transshipment terminals. First, metaheuristics such as local search solve the problem of scheduling incoming trucks as well as allocations at the loading gates taking into account the characteristics of the internal sorting system. Next, detailed conclusions on the real system behavior are drawn by testing the solutions in DES with stochastic processing times. A framework is used to create multiple simulation experiments for each solution. Results are investigated using the framework in order to identify the most robust solution provided by local search.

This paper introduces a practical approach for the comprehensive simulation-based planning and optimization of the production and logistics of a discrete goods manufacturer. Although simulation and optimization are established planning aides in production and logistics, the actual application in the field is still scarce, especially in SMEs – this is largely due to the complexity of the planning task and lack of practically applicable approaches for real life planning scenarios. This paper aims to provide a case study from the food industry, featuring a comprehensive simulation based planning. The approach utilizes an offline coupled multilevel simulation to smooth production and logistics planning via optimization, optimally configure the production system using discrete event simulation and optimize the logistics network utilizing an agent based simulation. The connected simulation and optimization modules are able to enhance the production logistics significantly, potentially providing a reference approach for similar industry applications.

Supply chain (SC) planning in the semiconductor industry is challenged by high uncertainties on the demand side as well as a complex manufacturing process with non-deterministic failure modes on the production side. Understanding the complex interdependencies and processes of a SC is essential to realize opportunities and mitigate risks. However, this understanding is not easy to achieve due to the complexity of the processes and the non-deterministic human behavior determining SC planning performance. Our paper argues for an agent-based approach to understand and improve SC planning processes using an industry example. We give an overview of current work and elaborate on the need for integrating human behavior into the models. Overall, we conclude that agent-based simulation is a valuable method to identify favorable and unfavorable conditions for successful planning.

We consider elements of a sustainable and distributed generation system for a wafer fab. Wind turbines (WTs), solar photovoltaics (PVs), a substation with grid access, and a net metering system are included in the generation system. WTs and solar PVs have the highest priority in supplying electricity. Surplus energy can be returned to the main grid. The objective function of the production planning formulation contains production-related costs, cost for energy from the substation, and penalty costs when a renewable energy penetration is not reached. This cost can be reduced by offering surplus energy to the main grid. The production plans are executed in a simulation environment to compute the profit in the face of machine breakdowns, wind power volatility, and uncertain power output of the PVs. The approach allows determining an appropriate number of WTs and solar PVs for demand scenarios. We present results of simulation experiments with the model.

Although production planning models using nonlinear CFs have shown promising results for semiconductor wafer fabrication facilities, the lack of an effective methodology for estimating the CFs is a significant obstacle to their implementation. Current practice focuses on developing point estimates using least-squares regression approaches. This paper compares the performance of a production planning model using a multi-dimensional CF and its robust counterpart under several experimental settings. As expected, as the level of uncertainty is increased, the resulting production plan deviates from the optimal solution of the deterministic model. On the other hand, production plans found using the robust counterpart are less vulnerable to parameter estimation errors.

In semiconductor manufacturing, new technologies impose more and more time constraints in product routes, i.e. a maximum time between two (often non-consecutive) operations. The management of Time Constraint Tunnels (TCTs, combining multiple time constraints) in high-mix facilities is becoming more and more challenging. This paper first recalls an approach for estimating the probability that a lot at the entrance of a TCT will leave the TCT on time. This approach relies on a list scheduling algorithm using a dispatching policy with random components. Three dispatching policies are presented. Computational experiments on industrial data comparing the three dispatching policies are presented and discussed. Perspectives are drawn to extend the approach and support decision making.

This paper is to achieve the on-time delivery and throughput for a semiconductor wafer fabrication (FAB) with dedication constraint. To be successful in the globalized competition, most of the companies have applied a concept of the production target which is defined as the quantity of production to be achieved for each day. Also, it is necessary to consider natural bias that significantly affects the alignment of patterns between different photolithography steps. As the natural bias has a negative effect on the quality and yield of products, most manufacturers have applied dedication constraint. To overcome the problem, we propose the two boundary based dispatching rule with the dedication constraint. The simulation model based on MIMAC6 was developed to prove the performance of this proposed dispatching rule, and conducted a simulation by using MOZART®. The simulation results clearly show the advantages of the proposed dispatching rule over the other dispatching rules.

In this paper, we discuss the construction of dispatching rules for semiconductor wafer fabrication facili-ties (wafer fabs) that take equipment health issues into account. Monitoring the equipment health status of critical machines is important to maintain process quality and to reduce rework and scrap. Usually, there is a tradeoff between quality- and productivity-related goals in wafer fabs. Obtaining an appropriate com-promise between these two objectives is addressed by considering blended dispatching rules. Simulation-based optimization based on variable neighborhood search (VNS) using a reduced simulation model of a wafer fab is applied to determine appropriate weights for the different priority indices. We demonstrate by simulation experiments that the obtained blended dispatching rule performs well compared to dispatching rules that are designed only for a single quality- or productivity-related objective.

Even in today’s semiconductor Fab the Automated Material Handling System (AMHS) is still the part one pays too little attention to. With having in mind that all the added value is done on the process tools this opinion might be comprehensible. But without the Overhead Hoist Transport (OHT) vehicles to be at the right time at the right place semiconductor manufacturing could not work efficiently. The following paper describes new empty vehicle management strategies as an important part of this on time AMHS delivery. Information about future upcoming transport jobs will be included to allocate this limited resource proactively and to achieve goals such as minimizing the tool waiting time for empty vehicles or the total number of dispatch moves. Different scenarios with changing input parameters will be tested and compared by using a simulation model which was developed with the focus of representing empty vehicle balancing functionalities.

Automated material handling systems (AMHS) can greatly impact the manufacturing performance of a semiconductor fabricating facility (fab). High traffic loads within an AMHS can impede individual wafer lots so that they arrive late at their destination machines. Thus, corresponding process operations as well as dependent succeeding operations will be delayed due to the fab schedule’s precedence constraints. Consequently, such transport-related delays can widely propagate throughout the overall fab schedule. In order to reduce transport-related delays before time-critical operations, novel ways of planning wafer transports have been investigated in this study. For validation, a well-known realistic representative wafer fab model has been extended with conveyor elements constituting a typical AMHS for continuous flow transport (CFT). As a result, improvements of the overall fab performance due to advanced transport scheduling methods are demonstrated and compared. Finally, the practicality of the suggested methods is discussed in the dynamic scheduling context of real fabs.

This paper addresses the design of an Automated Material Handling System (AMHS) for wafer lots in the photolithography workshop of a 200mm wafer manufacturing facility (fab) that was not initially built to have such a system. Lots transportation has to be performed using an Overhead Hoist Transport (OHT) system that was already chosen to transport reticles in the workshop. The main objective is to propose a decision support tool to characterize the AMHS elements including lot handling, transportation as well as the storage space design. A simulation-based approach is proposed to evaluate different scenarios and propose an effective AMHS design. Experimental results based on real instances confirm the capability of the proposed AMHS design to support the workshop activity.

We propose the creation of a new set of fab simulation testbeds. Extensions and additional features, not considered in the original MIMAC datasets, shall be incorporated in these new testbeds, thus allowing researchers to evaluate new methodologies with the same frame of reference. To do this, we surveyed the literature and mapped the pertinent research efforts of the past two decades. In this paper, we discuss in detail the various aspects of the new testbeds, in order to receive feedback from the simulation community on the importance of inclusion of some of the items in question; and the verification of the required inclusion of other items. Given the feedback, we aim to generate these testbeds within a year to serve as the new frame of reference for the benefit of the entire semiconductor manufacturing simulation community.

Semiconductor fabs are among the world’s most expensive factories, and present some of the most interesting manufacturing planning and control problems. It’s not a surprise that simulation is the workhorse analysis methodology, both in practice and in research. Over the past thirty years, a number of “standard” wafer fab test problems have been published to provide a common basis for comparing proposed planning and control policies and algorithms. We present a reference model, expressed in an analysis application agnostic language, OMG SysML™, that can be used to specify any of these test problems. We then show how this reference model also can be used to automate the generation of analysis models for one particular simulation solver, and argue for the use of parametric problem generators as a way to more fully explore planning and control methods.

The operational decision making in the BOSCH’s 200mm wafer fabrication facility has been guided by short term simulation forecasts. The forecasts provides the capability of identifying daily bottlenecks, forecasting daily fab outs, optimizing the preventive maintenance plans and personal resource planning. Now there is a pressing need to extend the forecast time horizon to several months for making decisions such as analyzing different ramp up scenarios, evaluating the impact of dispatch rules, identifying bottlenecks for capital investment, etc. As the short term model has achieved forecast accuracy of above 90%, it is used as the basis to generate the long term model. In this paper, we discuss the key issues associated with this model generation process. These issues are: process flows compression, flexible equipment dedications, model warm-up, wafer start generation, and future fab capacity changes. Our approach enables us to use the same model generation framework for both models.

In this work, we demonstrate that automatically classifying defects in topography images of silicon wafers is feasible. We process topography images of a set of sample wafers with controlled induced defects in their wafer back surfaces. We group these induced defects into three classes: cavities, cracks, and star cracks. With this sample set, we train and test selected classifiers with suitable feature vectors extracted from their wafer back surface topography images. A comparison reveals, that training and testing linear and quadratic classifiers with two Fisher scores as features, yield the best classification performances. We correctly classify all cavities and can separate them from the critical cracks and star cracks, which show a sufficient signal in the topography images.

Demand planning in the semiconductor industry is typically divided into different planning horizons, mid-term and short-term. Accurate demand forecasting is crucial because of long capacity installation times, long lead-times, short product life cycles, and constantly new technological advances. As demand forecasting for short and mid-term horizons are often made on different product and time granularities using different planning tools, we may see demand fluctuations (on the same granularity) within individual horizons and at the intersections of different granularities. This paper discusses stability of demand forecasts depending on time and product granularity and introduces definitions of good and bad stability, using Symmetric Mean Absolute Percentage Error (SMAPE) as a measure for stability. We show that time and product granularities have a significant effect on the intra-horizon stability of a demand plan and that planning on different granularities can lead to artificial demand fluctuations at the intersections of planning horizons.

Operations meeting in a wafer fab involves daily alignment of action items among key stakeholders: operations, maintenance, engineering and planning department. They have conflicting job functions. The maintenance department is required to conduct preventive maintenance (PM) to improve tools’ reliability The engineering department is required to qualify new products and processes. Both cases interfere with the flow of production lots. The planning department must ensure production ramp up. This can have a short term impact on overall fab delivery and capacity. The primary challenge is to reach aligned decisions: e.g. the best timing for PM or optimizing dispatch prioritization of production and development lots to ensure on-time delivery while maximizing capacity and tool utilization. In this paper we discuss the associated modelling issues of a 7-day simulation-based forecast, providing forecast of incoming WIP, moves and utilization at work center level. The simulation forecast consistently achieved an accuracy above 90%.

With the rapid development of semiconductor manufacturing, customers’ demand for on-time delivery rate (ODR) makes scheduling strategies face new challenges. In order to meet customers’ delivery requirements, scheduling strategies generally need to comprehensively consider remaining cycle time (CT), ODR, movement (MOV) speed and machine load balancing. In order to solve these problems, this paper proposed a scheduling strategy of semiconductor production lines with remaining cycle time prediction. Firstly, gather features related to performance index and then filtrate them through dimension reduction method. Secondly, use the above feature subset to build remaining cycle time prediction model by random forest algorithm. Next, design the scheduling strategy of semiconductor production lines with remaining cycle time prediction. Finally, make simulation experiments to verify the effectiveness of the proposed scheduling strategy. Simulation results show that the proposed scheduling strategy can improve the mean CT, throughput (TH), machine utilization time (MUT) and ODR in different extant.

This paper shows how automatic optimization of scheduling problems has been integrated within the automation framework of a semiconductor factory. Special attention is paid to the requirements arising from such an application in real world production in terms of constraints and objectives as well as from a factory integration perspective such as autonomous operation, high availability and efficient maintenance. Subsequently, possible solutions on how such requirements can be addressed will be discussed. Thereby, the advantage of using Constraint Programming solvers is highlighted. Knowledge that was gained during the implementation is presented for selected cases followed by the benefits that were achieved.

I had the opportunity to work for about 14 years on many different projects with two manufacturing sites of the French-Italian semiconductor company STMicroelectronics. Supported by European, national and industrial projects, this still active long-term academic-industrial collaboration led to many scientific and industrial achievements, spreading to other companies. Through regular exchanges, engineers, researchers, PhD and Master students were able to present their problems, their advances and generate new research projects. After some history of the collaboration, the presentation will survey some of the main research and industrial results in qualification and flexibility management, production and capacity planning, scheduling, automated transportation, dynamic sampling and time constraint management. Challenges faced and lessons learned when applying Operations Research and Industrial Engineering in practice, and in particular in semiconductor manufacturing, will be discussed. Benefits for both practitioners and researchers will be emphasized, such as the opportunity to propose and study new relevant problems and develop and apply novel approaches using actual industrial data.

The processes of implanting tools in semiconductor production involve a variety of setup parameters, suggesting utilization of a scheduler for planning lot allocation in a reasonable manner. However, a scheduler is complex by design, making evaluation difficult without appropriate aids. We describe an approach using discrete event simulation software for assessing the performance and supporting the parameterization of an implant scheduler. Furthermore, rolling-horizon approaches are discussed and the scheduler performance is compared to two dispatch rules.

This research is motivated by the need to verify and implement a schedule in a real production environment, especially in precarious production environments. This paper presents a mixed integer program (MIP) with time constraints and analysis risk parameters for tool interruptions. With the assistance of the survival analysis, a safety value will be computed and included in the MIP to downscale the available capacity. To verify the quality and robustness of the MIP, it is necessary to simulate tool interruptions and to change assumed release dates of production-bound jobs which have different stochastic distributions. To simulate these instabilities a hybrid model has been created which combines a discrete event simulation with a MIP solver. Finally, the results of the various simulations are compared.

The idea of deploying unmanned aerial vehicles, also known as drones, for delivery in logistics operations has inspired this research. One conceivable scenario is to use a drone to transfer jobs between locations in a future semiconductor factory. Each job might be characterized by origin, destination, priority, and precedence-relationship. In particular, the precedence-relationship occurs when drones are competing for limited number of ports (similar to helicopter landing platform). The objective is to minimize the maximum completion time of all delivery jobs performed by a fleet of drones. Two exact approaches are presented: a mixed integer programming and a constraint programming, and tested for real-time perspective with problem instances up to 50-drone and 100-job.

Quality is an important measurement within a semiconductor manufactory. Due to the fact that yield is directly affected by quality of the manufacturing process, in this paper a quality based scheduling approach will be presented which compares different methods like dispatching, MIP and CP, regarding different objectives. To test the different used methods a benchmark model of a semiconductor manufactory is build up. Here a lithography work center is used in detail where the rest of the fabrication is only build up as a delay station. With this model the repeatability for the example of a lithography step is investigated. Thereby in this investigation it is assumed, that each lithography tool has an offset which is transferred to the structure. Now the quality of a product should be best, if the offset from one layer to the next layer is minimized.

This paper proposes rescheduling algorithms for improving schedules obtained by dispatching rules with a commercial software program, MozArt, developed by VMS Solutions Co., Ltd.. Schedules for flexible flow shops with sequence-dependent setup times and job splitting are analyzed. The objective of the algorithms is to reduce the completion time by decreasing the number of setups and setup times. We first identify four types of problems with badly assigned job sequences and unnecessary idle times in given schedules derived from dispatching rules, and solve the problems by changing the sequence of jobs or splitting jobs. The performance of the proposed algorithms is tested with randomly generated instances based on real data from a factory in Korea.

In this study, we design and develop module-based modeling scheme that can conveniently represent a manufacturing system adopting a dual-card kanban system with a delivery cycle. By combining the designed modules, it is possible to express models of various systems comprising multiple parts, production lines, suppliers, and Mizusumashi (fixed-course pick-up). The module-based modeling scheme is useful for understanding the characteristics of just-in-time manufacturing and helping decision makers build simulation models based on the modules. We focus on modularization of the assembly line and the parts carrying Mizusumashi in a JIT manufacturing system. The proposed modules have focused dialogs, animation, and modeling functionality. In addition, a procedure for finding the optimal number of tray containers and kanbans to achieve no stock-out events is proposed based on the simulation model that integrates the kanban system and the Mizusumashi system. Then, the proposed procedure is applied to a numerical example.

Safety stock calculations are difficult for products with intermittent demand, long production lead times, and high monetary values. Theoretically, forecasts can be used to reduce the need for safety stocks. A high precision forecast minimizes the need for safety stock and forecast evaluation measurements can be used to calculate the safety stock level. However, a more realistic determination of safety stock levels can be obtained by simulation. In this paper, simulation is used to model and experiment on a case with three end products in order to determine the relationship between safety stock levels and service levels. Also, a comparison is made with theoretically calculated safety stocks to see how well basic theoretical models for safety stock calculations fulfill the requirements of service level. The result is that simulation can provide a much more accurate determination of safety stock levels for intermittent demands than theoretical calculations.

In micro manufacturing, the determination and scheduling of maintenance activities can strongly impact the production efficiency of the corresponding production system. Thereby, the supervision of quality relevant parts, tools and components can be very complex, due to the limited spaces within the manufacturing devices. This article proposes an extension of quality control charts by adding predictive component. This component predicts at which point in time maintenance activities are required based on quality characteristics of the produced work pieces. The article further presents two simulation studies. These demonstrate that the extended approach can compete with will configured time-based maintenance strategies in terms of production efficiency and rejection rates. In addition the predictive nature of this extension can issue forewarning for tool wear induced quality defects very early during production, allowing for an suitable integration of maintenance activities into the production schedule.

This work presents a simulative approach to combine material flow simulation with the energy flow simulation for a foundry use case. The casting process is strongly dependent on the conventionally true value of the temperature, and hence a high accuracy in the thermodynamic models is needed. The presented approach reaches this accuracy via the mathematical tool MATLAB. The coupled simulation of Plant Simulation and MATLAB models provides a possibility to combine a material flow simulation with a mathematical software. With the help of this tool even complex thermodynamic problems can be solved. Furthermore, this work examines different simulation scenarios and their results on the output and energy consumption of a foundry. The results show the influence of the charging intervals and the melt temperature on machine failure and hence on the casting part throughput.

Call centers have been the subject of many simulation studies and the challenges and successes of modelling these environments have been widely published. Call centers have however evolved significantly in the last decade or so, as new communication technologies have become available, customer expectations have increased, and the value of customer experience has been recognized and prioritized. While the single-channel call centers proved complex to analyze and worthy of the advanced analytical power of simulation, the modern multi-channel connection centers bring new levels of complexity to managerial decision making. This paper presents research into the modelling challenges posed by these evolving environments and illustrates how simulation is now more beneficial than ever to organizations aiming to understand and quantify the impact of change in their customer support business.

In the poultry processing industry demand and supply are still growing in volume and diversity, which requires more processing capacity, flexibility and smarter control. This paper focuses on the fillet batching process. To minimize the giveaway of fixed-weight fillet batching the right choices on layout, buffer sizes, batch sizes and batch allocation policies are of great importance. We develop a simulation model to support such decisions on design and control. The model is used (i) to determine buffer and grader sizes, (ii) to optimize batch allocation in a dedicated layout, (iii) to compare a dedicated to a flexible layout and (iv) to assess the impact of smart allocation policies. In particular we find that significant reductions in giveaway can be achieved by employing so-called index policies in a flexible layout.

Nowadays, due to the need of innovation and adaptation for the mass production of customized goods, many industries are struggling to compete with the manufacturing sector emerging in different countries around the world. The understanding and implementation of different improvement techniques is necessary in order to take part in the so-called fourth industrial revolution, Industry 4.0. This paper investigates how two well-known improvement approaches, namely lean and simulation-based optimization, can be combined with the concepts of Industry 4.0 to improve efficiency and avoid moving production to other countries. Going through an industrial case study, the paper discusses how such a combination could be carried out and how the different strengths of the three approaches can be utilized together. The case study focuses on how the efficiency of a production site can be increased and how Industry 4.0 can support the improvement of the internal logistics on the shop floor.

Selection of process improvement initiatives can be a challenging task. Process improvement projects usually fall into the following categories: Lean, Six Sigma, Lean Six Sigma, Change Management, and Business Process Reengineering. The selection process of these projects is a multi-criteria decision making process which involves multiple conflicting objectives. In this study, we develop an optimization model to select process improvement projects taking into consideration resource availability, required skills, and budget constraints. In addition, discrete event simulation (DES) models are developed to evaluate some of the selected projects. The DES models account for the uncertainty in the system and allow for performing scenario analysis on the selected projects. To validate the proposed approach, we provide a case study from a high-end server manufacturing environment. Results can be used to enhance the decisions on selecting process improvement projects.

Due to the increasing complexity of contemporary production scheduling problems, it is generally not possible to calculate nearly optimal production schedules in an acceptable amount of time. Hence, normally, dispatching rules are used to determine the job sequences. However, the selection of suitable dispatching rules is not a trivial task and depends on the relevant key performance indicators. Moreover, the suitability of dispatching rules changes over time because of the stochastic and dynamic nature of manufacturing systems. This paper proposes an adaptive simulation-based optimization approach to select individual dispatching rules for production control. The paper’s contribution is two-fold. First, it shows that the proposed approach improves the performance compared to benchmark approaches in a manufacturing scenario from semiconductor industry. Second, in order to be able to react quickly to dynamic changes, it proposes strategies for maintaining information from previously calculated solutions after a change, such as a machine breakdown, occurred.

Shipbuilding is one of the most complex manufacturing processes due to the high number and diversity of elements involved throughout the production process. Particularly challenging is the production of singular vessels, such as frigates, where effective planning becomes crucial for delivering the vessel on schedule due to the uniqueness of the product and the lack of historical data of previous equivalent constructions associated to this singularity. In this study, an ongoing simulation-based model that minimizes the uncertainties of the shipbuilding process is presented. Using the Discrete Event Simulation software ExtendSim, three real case studies are presented for the model validation. The objective is to obtain a multi-level model that can be used not only at early stages of the project, when detailed specifications are yet unavailable, but also at later stages, when design is well advanced and extensive data become accessible.

In this paper a practical case from an Austrian mechanical engineering company is presented. Simulation based manufacturing system improvement is applied to their component manufacturing plant. Based on the high number of items in the real case, a method for reduction of simulation model complexity applying item aggregation is developed in this paper. In the first improvement step, strategic capacity investment decisions are supported with the use of simulation. In the second step, a MRP planning parameter optimization is performed to improve service level and inventory. Additionally, the effect of capacity related decisions concerning setup time reduction and load-dependent outsourcing is evaluated. The results of this simulation study show that service level and inventory can be significantly improved by optimization of planning parameters and reduction of setup times. In addition, the study shows that load-dependent outsourcing is a viable alternative to capacity investment.

Capacity planning is a crucial task for economically sound production. Especially in semiconductor manufacturing, as equipment is expensive and production complex. An important part of valid capacity planning is a good understanding of equipment capabilities and characteristics and their influence on the workflow. Traditional approaches require new analysis with changing situations in the fab, which require special expertise and time. To enable a companywide standard and provide an easy to use tool, we are developing a utilization limit estimation tool. In this paper, we present our approach for a utilization target estimation system which bases its estimation on a wide range of data points created by data farming. Then, we apply a regression analysis to interpolate missing data points in order to provide fast estimates for utilization limits depending on equipment characteristics.

Real-time job shop scheduling is a sequential decision making problem. The main task is to decide which job in a queue should be processed. The problem can be modeled as a Markov decision process. Jobs in the queue form an action set. Selecting one job to process is regarded as taking an action from the set. A dummy action, which means no job will be selected and the machine will keep idle, is also contained in the action set. This removes the no-delay restriction from the problem. The reward function comprises the critical ratio of the selected job and the global job holding cost. Two algorithms, simulation-based value iteration and simulation-based Q-learning, are introduced to solve the scheduling problem from the perspective of a Markov decision process. The simulation explores the state space and accomplishes state transitions. The value function is parameterized and estimated by using a feedforward neural network.

The Industry 4.0 environment enables direct communication between the manufacturer’s shop floor and a customer. Thus, the manufacturer is able to respond to the customers’ requests more quickly, meaning that manufacturers must now more tightly control the shop floor planning and scheduling. Here we present a simulation-based scheduling model for Flexible Manufacturing System dynamic shop-floor control. The customer’s order and the processing sequence table of the products are imported into the simulation model. Experiments are implemented for the case wherein the system encounters unexpected conditions. The proposed approach represents a potential tool for manufacturers to make decisions in the real time by further connecting to the Enterprise Resource Planning and Manufacturing Execution System.

Production Planning and Control (PP&C) has been deeply analyzed in the literature, both in general terms and focusing on specific industries, such as the fashion one. The paper aims to add a contribution in this field presenting an optimization model for the Fashion Supply Chain (FSC), developed considering an interdependent environment composed by a group of focal companies that work with both exclusive and not-exclusive suppliers. The proposed framework will combine simulation and optimization models based on parameters, decision variables, constraints and Objective Functions (OFs) collected through a literature review. The framework has been developed in a parametrical way, in order to fit the peculiarities of the different actors operating along the FSC. The empirical implementation of the framework has been conducted using data coming from fashion companies belonged to the same network, considering rush orders as stochastics events for the scenario analysis and Key Performance Indicators (KPIs) assessment.

Simulation is a powerful tool to analyze and help in the decision making process of a production system. It is capable of delivering a dynamic analysis, both of the existing system and the future planned system. One major challenge with simulation projects however, is the time required at the initial stage when collecting data. For this study, Value Stream Mapping (VSM) has been selected as a complementary method for the data collection. VSM has been widely spread in industry, and it is a suitable method for identifying value streams and visualizing flows. In this study, the applicability of VSM for data collection is examined for a production system with complex and non-linear flows. The results of this study confirms that VSM can support in the data collection phase, but entails the support from subject matters.

Defect management in manufacturing environments requires effective identification of the defects and finding proper solutions to resolve them. Predicting and preventing the defects before they can occur is the focus of quality risk management. To effectively manage defects, companies need to analyze historical data to identify the causes and solutions for defects as well as study the impact the defect can have on the processes, priorities, and operations. This study integrates data analytics and simulation modeling to develop a system for defect management in manufacturing environments. Simulation is used to analyze the behavior of the system whereas data analytics is used to develop prediction models for defect resolution. A case study from high-end server manufacturing environment, which is characterized by extensive test processes to ensure high quality and reliability of servers, is provided. The proposed approach helps decision makers analyze and manage defects and develop proactive means to prevent them.

Discrete event simulation is an established methodology for investigating the dynamic behavior of complex manufacturing and logistics systems. Traditionally, simulation experts conduct experiments for predetermined system specifications focusing on single model aspects and specific analysis questions. In addition to that, the concept of data farming and knowledge discovery is an ongoing research issue that consists of broad scale experimentation and data mining assisted analysis of massive simulation output data. As an extension to this approach, we propose a concept for investigating the robustness of complex manufacturing and logistic systems which are often very sensitive to variation and noise. Based on Taguchi’s loss function, we developed a concept including data farming and visual analytics methodologies to investigate sources of variation in a model and the factor values that make a configuration robust. The concept is demonstrated on an exemplary case study model.

Digitalization is an ongoing revolution within manufacturing industry. 5G technology is expected to play an important role in ensuring connectivity. Digitalized factories set high requirements on technical availability, and therefore also on maintenance performance. However, it is difficult to get top-level decision makers to invest in maintenance, since the effects are usually deferred and difficult to verify up front. For quantifying long term effects, Discrete Event Simulation (DES) is identified as a powerful tool. In this study, DES was combined with established maintenance concepts to provide analysis of a real-world industrial 5G pilot implementation. Maintenance concepts were used to identify relevant inputs and outputs to the simulation model. The model was tested on a use case, where 5G enables support for maintenance tasks. By applying DES and maintenance concepts on more use cases, there is a potential to quantify effects of maintenance and enable digitalized production in a larger scale.

Factory layout planning is essential for manufacturing companies when designing or redesigning production systems. Layout planning usually involves 2D CAD applications, sometimes based on faulty data. Difficulties in communicating and discussing layout alternatives using such applications can lead to critical errors, yielding inaccurate simulation models producing poor results. This paper aims to investigate and evaluate the usefulness of realistic 3D layout models in the layout planning process, addressed by an industrial study of how existing methods for visualization can be applied. This paper shows that utilizing a realistic and accurate layout model allows for fruitful discussions while several potential mistakes can be avoided. It also shows benefits in evaluating a layout and the model’s accuracy in immersive virtual reality where a better perspective of the layout can be acquired. Having such evaluated layout models will enable more accurate simulation models for planned changes, based on real physical requirements.

Although the idea of using Augmented Reality and simulation within manufacturing is not a new one, the improvement of hardware enhances the emergence of new areas. For manufacturing organizations, simulation is an important tool used to analyze and understand their manufacturing systems; however, simulation models can be complex. Nonetheless, using Augmented Reality to display the simulation results and analysis can increase the understanding of the model and the modeled system. This paper introduces a decision support system, IDSS-AR, which uses simulation and Augmented Reality to show a simulation model in 3D. The decision support system uses Microsoft HoloLens, which is a head-worn hardware for Augmented Reality. A prototype of IDSS-AR has been evaluated with a simulation model depicting a real manufacturing system on which a bottleneck detection method has been applied. The bottleneck information is shown on the simulation model, increasing the possibility of realizing interactions between the bottlenecks.

There are many projects using discrete event simulation as a decision-making tool. However, it was found in the literature that these articles are related to improving the execution of projects, not studying issues related to the management of simulation projects. From this perspective, this paper used concepts proposed by the PMBOK® to drive the management of communication between members of a simulation project and stakeholders. To achieve this goal, a real simulation case in a manufacturing company was studied. The steps of communication structure were followed and an Action Plan was elaborated and applied in this project. One of the outcomes of this effort is a communication model that was proposed for simulation projects. A questionnaire was applied in order to evaluate the proposal and results showed that simulation analysts judge communication aspects in their projects to be very important, and is strongly related to the success of the projects.

Discrete Event Simulation (DES) project data management is a complex and important engineering activity which impacts on the organization efficiency. This efficiency could be decreased due to the lack of provenance information or to the unreliability of the existing information regarding preceding simulation projects, all of which complicates the reusability of the existing data. This study presents an analysis on the management of simulation projects and their provenance data according to the different types of scenarios usually faced in a manufacturing plant. A survey based on simulation projects from an automotive manufacturing plant was undertaken to categorize the information regarding to the studied projects, map the contained provenance data and standardize its management. This study also introduces an approach on how a structured framework based on the specific data involved in the different types of scenarios could allow an improvement on the management of DES projects.

Mobility modeling is a critical aspect of the ground vehicle acquisition process. Mobility modeling for traditional ground vehicles is well-understood; however, mobility modeling tools for evaluating autonomous mobility are sparse. Users do not understand the performance capabilities of autonomous ground vehicles at a mission level because no mission-level mobility model exists for autonomous vehicles. Therefore, this paper proposes a Reference Autonomous Mobility Model (RAMM). The RAMM serves as the mission-level mobility modeling tool that is currently lacking in the UGV community. The RAMM is built on the framework already established by trusted mobility modeling tools to fill the current analysis gap in the autonomous vehicle acquisition cycle. This paper gives a detailed description of the RAMM along with an example applications of the RAMM for modeling autonomous mobility. Once fully developed, the RAMM could serve as an integral part in the development, testing and evaluation, and fielding of autonomous UGVs.

The United States Coast Guard (USCG) plans the world-wide operations of its ships, referred to as cutters. Rule-based decision making mechanisms, the primary planning tools, are aimed at ensuring adequate time inport for maintenance, stand-by, and generalized assignments for at-sea deployments to geographic regions. This approach is problematic. Stochastic events are becoming increasingly costly. In 2013, Coast Guard cutters suffered nearly four times the normally allocated amount of operational days lost to unplanned repairs. Additionally, the predominant metric used by schedulers is the amount of time a cutter is focused on a single, primary mission. The reality of multi-mission operations is that unplanned emergency missions often supersede planned, routine missions. Furthermore, mission results contribute to multiple strategic goals, though to varying extents based on geographic area. We propose a comprehensive multi-mission schedule evaluation mechanism that measures productivity, given limited Coast Guard resources, across a spectrum of current and future scenarios.

Using anti-air warfare tactics and concepts of operations, we explore the dominant factors for convoy operations. A discrete event simulation that facilitates modern analysis was built to model ships, their sensors, and their weapons. The model was used to simulate over 1.5 million naval battles in which we varied 99 input variables using a nearly orthogonal nearly balanced (NOB) Latin hypercube design of experiments. Metamodels were then constructed to study what impact the factors have on the survival of a High Value Unit, and give guidance for which factors offer the most improvement. In addition to our specific findings, this study can be used as a guide for how to conduct future analyses.

A discrete event simulation model is developed to represent a forced distribution performance appraisal system, incorporating the structure, system dynamics, and human behavior associated with such systems. The aim of this study is to analyze human behavior and explore a method for model validation that captures the role of subordinate seniority in the evaluation process. This study includes simulation experiments that map black-box functions representing human behavior to simulation outputs. The effectiveness of each behavior function is based on a multi-objective response function that is a sum of squared error function measuring the difference between model outputs and historical data. The results of the experiments demonstrate the utility of applying simulation optimization techniques to the model validation phase of simulation system design.

This paper deals with manpower planning using a dynamic and interactive simulation system that is agile and adaptive to robustly accommodate change - without requiring a complete rewrite. The simulation architecture extends the current hybrid modelling paradigm, which integrates agent based (AB) constraints and controls, with a discrete event simulation (DES) methodology. This allows for a more expressive, authentic representation of both process flows and agent policies that captures the advantage of system dynamics (SD) modelling by integrating agile lever controls with response feedback. This approach is inspired by the need to develop an aircrew training pipeline simulation for the Australian Defence Force (ADF) that supports the real needs for strategic manpower planning in a context of policy and requirements change management. A case study is provided to illustrate the challenges and approach.

Most swarming algorithms require individual nodes to know the locations of their nearest-neighbor peers. Existing work assumes that this information is abundant and readily available, however, when wireless communications are used for data exchange, issues regarding dissemination arise. These include partial or complete data loss and an increased latency, significantly affecting the quality of the delivered service. In this paper, we show through extensive experimental analysis that swarm intelligent algorithms are vulnerable to degraded communication. To show how communication is affected in a contested environment, we introduce a local interaction statistic metric to capture emergence. Our analysis using agent-based simulation characterizes the decay of inherent emergence and swarm efficiency with increasing data loss and delay.

We propose a new local networking metric, the network layer connectivity awareness (NLCA), to dynamically characterize the connectivity status at the network layer of mobile ad hoc networks (MANETs). The NLCA is a local view of routable destinations provided by a designated routing protocol, which may differ from the real-time physical layer connectivity (PHYCON), defined as destinations that can be reached by local nodes via (multi-hop) radio links. Such discrepancy can cause packet delivery failure because a route may no longer be available at physical layer. We present a simulation method to obtain the real-time PHYCON using the breadth-first search algorithm. We apply the NCLA metric to the optimized link state routing (OLSR) protocol in scenarios simulating tactical MANETs, and compare the resulting NLCA with the underlying PHYCON, illustrating the two measurements vary and differ under mobility. The proposed NLCA and investigation technique provide a method for routing performance analysis.

Does the national security exception in international refugee law constitute a real, legally measurable justification, or rather an excuse for introducing the politics according to the will of current government? A lack of sufficiently comprehensive set of rules within the 1951 Convention Relating to the Status of Refugees entails large, but not unlimited, subjectivity in interpreting legal norms. This work briefly presents the relevant legal norms and proposes a model aiming to balance between refugees’ rights and national security interests. The model can help in limiting subjectivity during adjudication process by quantifying the boundaries to implementation of the limitations of refugees’ rights. The provided example demonstrates the use of the model.

The development of tools that appropriately simulate electric-natural gas interdependencies and their resulting propagation of disturbances within and between systems is a particular need of system operators. The need for such tools is further emphasized by the January 2014 Polar Vortex event, where the response of the electric power and natural gas systems further highlighted the importance of the coordinated assessment of interdependent systems when a large diversion of natural gas to non-electric customers created unexpected consequences in the electric sector. This paper documents ongoing modeling and assessment activities of the Argonne Electric Power-Natural Gas Integrated Model and demonstrates the estimated impacts for this historic disruptive event. This paper presents a description of the Polar Vortex event, the modeling approach and methods used, including the assumptions, data, and modeling platform, and provides simulation results showing agreement with historically reported impacts, in both spatial and quantitative terms, to validate model performance.

Our Training and Experimentation Cloud Architecture, namely the hTEC, is applied to joint military space operations simulation. The hTEC follows the recommendations on the modelling and simulation as a service (MSaaS) by NATO Science and Technology Organization. The space mission areas and their characteristics are investigated and the requirements are analyzed. The hTEC services are designed such that these requirements are fulfilled. Each service addresses the minimum set of functions that may be needed by a military space operations mission area, and can be run independently, federated as a composed service or linked into a software application. The designed joint military space operations simulation architecture is implemented in a testbed called the extended BSigma.

Automated planning is a key to unlocking the next generation of human behavior modeling for military simulations. Automated planning is distinguished from other dynamic behavior by its ability to reason with predictions or assumptions about future outcomes, which has traditionally been left to the meticulous effort of human modelers. In this paper, we present a conceptual planning framework as an architectural roadmap for the development of this kind of capability in support of modeling and simulation. This paper also presents an initial implementation of the framework in a representative combat simulation for exploring potential future opportunities for community advancement of the planning techniques.

The usefulness of software frameworks to support the development of military combat simulations is gaining attention. Using a framework increases model reuse and can avoid the duplication of infrastructure code used to support model and simulation application development. Simulation frameworks encourage defining abstractions for the domain of interest, which allow for multiple concrete (i.e., specific) implementations of models at varying levels of fidelity, resolution and/or detail to be produced and assembled. This flexibility leads to customized simulation applications that are focused and aligned to an envisioned conceptual model of a system of interest. However, the difference between a framework and a specific simulation application, and its relationship to experimentation is not always clear. This paper elaborates on these distinctions and addresses how software frameworks support experimental objectives.

Human simulation (applying Modeling and Simulation (M&S) to topics in the humanities, the interpretative social sciences, and the arts) is a potent extension of social simulation. This paper offers reflections on teaching at this intersection, presenting best practices in pedagogy for undergraduate and graduate students engaged in formal studies, and for established researchers having no structured curriculum. The fact that human simulation is possible drives home the presence of formal patterns in a host of phenomena that for a long time were thought to be inimical to mathematical analysis. That implies a double pedagogical challenge: teaching humanities students to recognize formal structures in the phenomena they study (counter-intuitive for them), and teaching M&S students to collaborate with humanities people who think very differently (equally counter-intuitive). The three perspectives presented here underline the usefulness of human simulation, as well as the difficulties and benefits associated with teaching and learning human simulation.

R is free software for statistical computing, providing a variety of statistical and graphical functionality. For use in simulation education, R's capabilities help to develop student intuition. In this paper, we introduce the "simEd" package for R, written with a pedagogical focus. The package includes functions for generating discrete and continuous variates via inversion, with capabilities for independent streams and antithetic variates; for visualizing inversion in variate generation and the relationship to the pdf/pmf, cdf, and ecdf; for computing time-persistent statistics; for extensible single- and multiple-server queueing simulation; and includes data sets for input modeling and analysis. As we demonstrate using several illustrations, this package, along with native R functionality, provides a compelling case for using R in an introductory simulation course.

In this article, a cross-tooling method for automated model verification is presented using a reference model. Furthermore, the method is implemented in teaching, using a web platform. The method bases on Yücesan and Schruben (1992), demonstrating a procedure for the examination of a structural and behavioral equivalence of two simulation models based on Simulation Graph models. However, Simulation Graph models are subject to an event-oriented modeling world view. Since current simulation tools use a process-oriented - easier to understand - modeling world view, a simple queuing model shows how transformation from a process-oriented world view (Simio, AnyLogic) takes place in an event-oriented world view (Simulation Graph models). A further step then checks the structural equivalence via an isomorphic mapping on the resulting planar graphs.

Simplification is considered a fundamental part of modelling and simulation. Model simplification is instrumental in creating models that are useful – by focusing on system elements that matter, and feasible – by reducing study efforts. Despite its widely acknowledged relevance simulation model simplification may still be considered an underdeveloped field. This is mirrored in existing literature, and course books. While the former shows a fragmented landscape in addressing the issue, the latter often offer little guidance for the (future) analyst. To foster further development of the field we assess current progress by providing a literature review. Issues addressed by the review concern: (i) definition and scope of model simplification, (ii) reasons for model simplification, (iii) drivers of inappropriate model complexity, and (iv) approaches for model simplification. The review is meant to provide a useful overview of the work undertaken in this field, aiming to benefit educators, practitioners and researchers.

In this paper, we raise the questions: how could sound influence the usability of simulations? How could sound influence the learning of simulation creation? How could sound support processes like verification? We argue that sound can support learning by relying on music/sound cues’ emotional engagement on users and verification by providing insight into the correctness of simulation execution during runtime. For instance, sound cues could indicate when certain events occur and if processes in a simulation are operating within their specifications. We explore potential benefits and challenges posed by incorporating sound into DES models. Many perceived challenges of this incorporation overlap with known visualization challenges for conveying information during runtime, as both cases deal with conveying sensory stimuli. We present conceptual examples and report on ongoing efforts to integrate sound into a DES simulation environment.

The present work discusses the use of NetLogo, an agent-based simulation software, as a tool to teach ethics modeling as part of an ethics course. It allows students to define and describe the behavioral rules of agents under different ethical theories through agent-based simulation and learn and assess the consequences of such ethical behaviors. Several simulations developed by the students along with their principal findings are presented.

Uncertainty in planning tasks such as processing times, set-up times, customer required lead times, due dates, time to failure, time to repair and the complexity in terms of product variety, outsourcing, short lead times, low inventory levels, low costs and high utilization are major hurdles for planning logistics and production processes. This paper introduces a simulation-based business game for methods in planning logistics and production processes. Basic methods such as material requirement planning, Constant work in progress (Conwip), Kanban, reorder policies, dispatching rules, basic demand forecasting methods and master production schedule (MPS) are implemented in the game. Due to the generic environment additional methods can be implemented efficiently. The midterm planning concept sales and operations planning (S&OP) is implemented as well, where the gamers have to act as managers responsible for purchasing, production, sales and finance. Their target is to identify sales and production volumes for the next planning periods.

The sim4edu.com project website supports web-based simulation with open source technologies for open science and education. It provides both simulation technologies and a library of educational simulation examples. Its aim is to support both the use and the development of various kinds of simulations, including ad-hoc simulations, Cellular Automata models, NetLogo-style grid space models, discrete event simulation and agent-based simulation. Sim4edu facilitates building state-of-the-art user interfaces for web-based simulations and simulation games without requiring simulation developers to learn all the recent web technologies involved (e.g., HTML5, CSS3, SVG and WebGL). Using the power of the web, Sim4edu allows researchers and educators to publish and share their models easily.

The 50th anniversary of WSC provides a good opportunity to reflect on how widely simulation is used in the business world to support problem solving and decision making. This paper posits that, despite the success of the WSC, simulation is not as widely used as it should be and that a major cause is a general lack of understanding of the value of simulation outside of the simulation-specialist community. However, this paper suggests that one way to increase the use of simulation is through changes in education. As part of the review and development of suggested changes, the paper examines the activities related to education at the WSC. It then offers a variety of suggestions for how the simulation education community – including academics, practitioners, and vendors – can help to address this issue.

This paper and the corresponding panel session focus on teaching undergraduate simulation courses. The format brings together four experienced instructors to discuss four questions involving the structure and topic outlines of courses, print and software teaching materials used, and general teaching methods/philosophy. The hope is to provide some experience-based teaching information for new and soon-to-be instructors and to generate discussion with the simulation education community.

One of the characteristics of being human is to model. In our history, we began with representations of animals made from natural materials, and painted on cave walls. We also made regular marks on animal bones. While the modern accounting of these products is art (animal representations) and mathematics (bone marks), a more comprehensive understanding points to modeling in both cases. Since the inception of modeling, we created areas of knowledge and have divided things into many groups. These groups have sub-groups to where our knowledge resembles a large house with its artificial partitions. And yet, modeling is still pervasive although it differs slightly in form among these subdivisions that we now refer to as disciplines. We claim that models are natural transformers from human experience to information; to create information for object X, create a model of X.

When learning to create simulations, we rely on real systems to emphasize their importance on reality. However, for younger students, reliance on reality is not always engaging. Reality provides context, but students’ interest quickly fades. Through the use of four case studies, we explore the idea of having students create stories in order to engage them in learning to create simulations. Applying a narrative/story context provides a mechanism for learning and maintaining student engagement. Stories can be based on original/existing games, movies, or other sources rich in narrative. Our approach includes four components: create a game/story narrative; discuss, evaluate, and expand the narrative; implement the narrative into an animated storyboard; and implement the narrative into a simulation. Lastly, we briefly discuss the utilization of sounds on both the animated storyboard and the simulation. Future work will empirically explore the effectiveness of narrative storytelling for learning simulation creation.

This paper presents a new educational roadmap for teaching discrete event system (DES) simulation software design. This roadmap represents the hierarchical structure and inter-relationships characterizing the worldviews of DES simulation, namely, event scheduling and process interaction. The roadmap was developed from the authors’ experience while teaching DES simulation to both undergraduate and graduate students, spanning several years. The roadmap’s development was motivated by the need to strive for greater completeness as well as fewer inconsistencies in the material curriculum. The commonality between the worldviews is highlighted in striving for a uniform approach. The intent of this paper is to provide other educators a foundation for their own DES simulation course development. A simple example is used to illustrate the worldviews within the roadmap.

Human simulation (applying Modeling and Simulation (M&S) to topics in the humanities, the interpretative social sciences, and the arts) is a potent extension of social simulation. This paper offers reflections on teaching at this intersection, presenting best practices in pedagogy for undergraduate and graduate students engaged in formal studies, and for established researchers having no structured curriculum. The fact that human simulation is possible drives home the presence of formal patterns in a host of phenomena that for a long time were thought to be inimical to mathematical analysis. That implies a double pedagogical challenge: teaching humanities students to recognize formal structures in the phenomena they study (counter-intuitive for them), and teaching M&S students to collaborate with humanities people who think very differently (equally counter-intuitive). The three perspectives presented here underline the usefulness of human simulation, as well as the difficulties and benefits associated with teaching and learning human simulation.

Cognitive biases such as myopic problem representation, group think, the conjunction fallacy, and confirmation bias impair effective decision making. Therefore, successful leadership should be able to spot and mitigate circumstances that promote cognitive biases. To understand and experience the consequences of cognitive biases and to teach and train on adequate countermeasures (debiasing techniques), this article presents a serious game for leadership practice. In a round-based game, a group of participants works as a team to make decisions constrained by time, uncertainty, lack of information, and conflict. The game simulates the last stages of finalizing deep sea oil exploration. While the game focuses on behavioral learning outcomes, the final debriefing emphasizes emotional aspects to achieve a long-lasting learning effect. The debriefing reveals that the game is not artificial but mimics exactly the stages that led to the real catastrophe at the Deepwater Horizon rig in the Gulf of Mexico in 2010.

This contribution presents a model for dynamic networks of physical contacts among large populations and their application for reproducing complex patterns in epidemic spread. The networks are constructed from statistical data on demography, geography, organizational structure and contact behavior. Due to the heterogeneous nature of the data and by construction, rich topological characteristics such as overlapping communities, layering along multiple dimensions and multiplex dynamics on different time scales can be observed. The generated dynamic networks can furthermore be regarded as subgraphs or derivatives of latent social networks. General results and observations form social network theory apply naturally and are used for explaining dynamic effects in epidemics. An exemplaric analysis investigates the impact of weak ties and effects of communities with decreased immunization on epidemic spread. Optimized implementation and visualization techniques turn out to be a key asset for dynamic simulation of contacts within large populations.

Human behaviour simulation models having cognitive, physiological and social dimensions of behaviour has been of interest for a long time. Within this landscape however work has lagged in 2 vital areas. There has not been much attention paid to models grounded in the behavioural sciences nor on techniques to create such models. In this work, we address both lacunae. We present a compositional approach to create grounded human behavioural simulation models and demonstrate its use with an example. The compositional approach has 3 components, a behavioural science element, a statistical and a computational. The behavioural component uses approaches to theory development in the behavioural sciences particularly in the realm of organizational behaviour. The statistical element derives valid chains of relations. Finally the computational element converts the base model into a simulate model. We demonstrate our approach by creating and executing a simulation model in the workplace domain.

The present study addresses the grouping-assignment problem of heterogeneous workers for serial and parallel manufacturing configurations considering the worker production rate as a function of learning by doing and knowledge transfer. A simulated experiment is presented for this end, considering the maximization of the system output as the optimization goal, and the system size and tasks heterogeneity as experimental factors. Three heuristic policies are compared based on the heterogeneity of the groups with respect to the individual knowledge transfer parameter. Research related to team formation for manufacturing systems is scarce and often does not considering workers’ heterogeneity nor the knowledge transfer. The results highlight the importance of considering workers’ heterogeneity for the grouping and assignment of workers. The implications of this study impact the managerial decision-making process related to the grouping and allocation of workers to tasks as part of the production planning.

With an increasing number of cybersecurity attacks due to insider threats, it is important to identify different attack mechanisms and quantify them to ease threat mitigation. We propose a discrete-event simulation model to study the impact of unintentional insider threats on the overall system security by representing time-varying human behavior using two parameters, user vulnerability and user interactions. In addition, the proposed approach determines the futuristic impact of such behavior on overall system health. We illustrate the ease of applying the proposed simulation model to explore several "what-if" analysis for an example enterprise system and derive the following useful insights, (i) user vulnerability has a bigger impact on overall system health compared to user interactions, (ii) the impact of user vulnerability depends on the system topology, and (ii) user interactions increases the overall system vulnerability due to the increase in the number of attack paths via credential leakage.

Within the archaeological record for Bronze Age Chinese culture, there continues to be a gap in our understanding of the sudden rise of the Erlitou State from the previous late Longshan chiefdoms. In order to examine this period, I develop and use an agent based model (ABM) to explore possible socio-politically relevant hypotheses for the gap between the demise of the late Longshan cultures and rise of the first state level society in East Asia. I test land use strategy making and collective action in response to drought and flooding scenarios, the two plausible environmental hazards at that time. The model results show cases of emergent behavior where an increase in social complexity could have been experienced if a catastrophic event occurred while the population was sufficiently prepared for a different catastrophe, suggesting a plausible lead for future research into determining the life of the time period.

The applicability of Wedding Ring model to an agent-based demographic module is investigated. As interpretation of social simulation results requires in-depth understanding of phenomena from various angles in the spatio-temporal dimension, the agent-based modeling (ABM) approach in demographic and social studies gets popular. To provide with quantitative information about chronological dynamics, computational cost and influential networks for the implemented model, we first introduce the Wedding Ring model which captures empirical marriage patterns based on a micro-level hypothesis and its outputs which are directly affected by social interactions. Then, the key observations are examined at different points in time, population size, and distribution of social networks to check its desirability for representational details of ABMs. In conclusion, the presented model could be extended into a large-scale social laboratory with acceptable predictability and computation load to allow for investigating non-numerical characteristics while supporting traditional approaches in a complementary rather than competitive way.

Civilian protest is a complex phenomenon where large numbers of protestors participate in demonstrations. It involves multiple groups, various trigger events and social reinforcement where groups excite each other. We present a graphical generative model in which a baseline spontaneous process may undergo excitation due to external triggers, as well as inter-group contagion. We define a trigger-conditional multivariate Hawkes process, where excitation is conditional on the presence of active triggers. An arrival in this process corresponds to a batch of protestors, and random marks on the arrival serve to capture both the excitation-related parameters as well as the size of protest. The batch arrival intensity and the batch size, while mutually independent, exhibit respective history-dependence due to memory that is modeled in the excitation phenomena. We present a simulation algorithm for generating sample paths, and results estimating likelihood of large-scale protest on a realistic model.

Arena 15.1 was released to the market earlier this year. This latest release of Arena includes a number of advancements designed to enhance ease-of-use, make complex simulations easier and expand the overall simulation capabilities of Arena. This presentation will cover those enhancements and provide demonstrations on how to make the most of these new capabilities.

AnyLogic Company produces the standard in multimethod modeling technology which equates to increased efficiency and less risk when tackling complex business challenges. This unmatched flexibility is found in all AnyLogic products allowing users to capture the complexity of virtually any system, at any level of detail, and gain deeper insight into the interdependent processes inside and around an organization.

Experience the latest business challenges solved using AnyLogic Simulation Software and anyLogistix Supply Chain Software in Cardinal Health, CSX, DHL, Deloitte, GE, GlaxoSmithKline and more.

This paper describes the Simio modeling system that is designed to simplify model building by promoting a modeling paradigm shift from the process orientation to an object orientation. Simio is a simulation modeling framework based on intelligent objects. The intelligent objects are built by modelers and then may be reused in multiple modeling projects. Although the Simio framework is focused on object-based modeling, it also supports a seamless use of multiple modeling paradigms including event, process, object, systems dynamics, agent-based modeling, and Risk-based Planning and scheduling (RPS).

Currently there is a lack of detailed understanding regarding the effect of autonomous vehicles on traffic operations and transportation infrastructure. PTV Vissim, the world‘s leading microscopic traffic simulation tool, provides a virtual testbed to evaluate the coexistence of autonomous and conventional vehicles either in the transition phase or when vehicle fleets are fully autonomous. Vissim traffic simulation is increasingly being employed to address the evidence gap around the potential impacts of disruptive technologies, such as connected autonomous vehicles, on traffic flow and capacity. Incorporating hardware in the loop testing allows detailed representation of autonomous vehicle control algorithms, sensor and communication protocols to be simulated within a realistic virtual traffic environment. Simulating varying penetration rates of autonomous vehicles, and different behavioral characteristics such as platooning, vehicle to vehicle communication and vehicle to infrastructure communication, has demonstrated the opportunity to reduce travel times by 11% and delays by more than 40%.

Organizations interested in applying simulation to support informed decision-making need to make a key decision whether to contract the work or develop an internal capability to perform the work. Both alternatives have its pros and cons and the ideal approach varies across organizations. This presentation reviews some of the key factors worth consideration, including the caliber of staff, costs, risks, and necessary and perceived internal focus. Furthermore, managers and developers have different responsibilities and challenges in ensuring a successful project, very much influenced by the structure of the simulation work being performed. The presenters will share experiences working both in an internal simulation role and working as external service providers.

Deploying simulation models to “business consumers” is now easier than ever. Analytic Solver® software offers a simple, point-and-click way to create and test analytic models using your web browser or your Excel spreadsheet – then, with just a few mouse clicks, make those models run in Tableau and Power BI dashboards. Your models can connect to any Tableau or Power BI data source, display results using charts and tables, and they’ll re-run whenever the data changes. If you’d rather deploy models on your own website, or in your mobile app, that is surprisingly easy as well. With powerful tools for Monte Carlo simulation and risk analysis, conventional and stochastic optimization, forecasting, data mining and text mining built in, you can use Analytic Solver to quickly build and deploy models that yield real business impact. We’ll demonstrate these capabilities in our Vendor Tutorial session at WSC 2017.

Discrete-event simulation as a methodology is often inextricably intertwined with many other forms of analytics. Source data often must be repaired or processed before being used (indirectly or directly) to characterize variation in a simulation model. Collection of simulated data needs to coordinate with and support the evaluation of performance metrics in the model. Or it might be necessary to integrate other analytics into a simulation model to capture particular complexities in the real world system. We show how SAS Simulation Studio, as an integral part of the SAS analytic ecosystem, enables you to tackle all of these challenges. You have full control over the use of input data and the creation of simulated data. Strong experimental design capabilities mean you can simulate for all needed scenarios. Additionally, you can embed any SAS analytic program—optimization, data mining, or otherwise—directly into the execution of your simulation model.

This presentation shows how to use MATLAB and its simulation tools Simulink and SimEvents to build hybrid simulations, such as a simulation for estimation production throughput and scheduling. Then, we show how to use MATLAB tools for statistics, machine learning, and optimization to understand the model, as well as use the model to make decisions. We show how to use Live Scripts and Apps for interactive lectures. We also show how to access MATLAB Online through a browser or MATLAB Mobile from your phone. Create and grade assignments with Cody Coursework.

Simulation has traditionally been applied in system design projects where the basic objective is to evaluate alternatives and predict and improve the long term system performance. In this role, simulation has become a standard business tool with many documented success stories. Beyond these traditional system design applications, simulation can also play a powerful role in scheduling by predicting and improving the short term performance of a system. In the manufacturing context, the major new trend is towards digitally connected factories that introduce a number of unique requirements which traditional simulation tools do not address. Simio has been designed from the ground up with a focus on both traditional applications as well as advanced scheduling, with the basic idea that a single Simio model can serve both purposes. In this paper we will focus on the application of Simio simulation in the Industry 4.0 environment.

Managers need state-of-the-art tools to help in planning, design, and operations. The AutoMod product suite from Applied Materials has been used on thousands of projects empowering engineers and managers to make the best decisions. AutoMod’s capability to model large complex automation and material handling systems continues to lead the market. Hierarchical model construction allowing users to reuse model components, decreasing the time required to build models. Recent enhancements to AutoMod’s material handling systems have increased modeling accuracy and ease-of-use. These advances have made AutoMod one of the most widely used simulation packages.

AutoMod’s power lies in its performance, scalability and accuracy in detailed modeling of large and complex manufacturing, distribution, automation and logistic operations, leaving the competition behind.

Come see the latest release and exciting plans for AutoMod’s future and why it has outlasted the competition.

Transparency is an essential attribute to meet the business objectives if the production is made under complex supply chain. At the same time, it is difficult to achieve if the production requires complex processes like semiconductor manufacturing or tons of sub-parts such as ship building and offshore industry. MOZART has been implemented in semiconductor, display panel, and tire industries as a planning and scheduling system. It covers weekly planning (Master Plan: MP), daily planning (Factory Plan: FP), and real time scheduling for these double-digit-day cycle time product manufacturing. MOZART extended the coverage to meet offshore project whose cycle time is several hundreds of days.