This tutorial gives an overview about the state-of-the-art of Digital Twins (DTs). We discuss what they are, what purposes they may fulfill, how they differ from related technologies, and how they may work internally. Application-wise, we focus on Digital Twins of production and logistics systems, but the majority of our explanations should also be valid beyond these domains. Finally, we discuss open issues and potential directions of future research.

Agent-based modeling and simulation (ABMS) has become one of the most popular simulation methods for scientific research and real-world applications. This tutorial paper explores recent development in the use of artificial intelligence including large-language models and machine learning, and digital twin in ABMS research. Given the different perspectives on ABMS, this paper will start with ABMS basic concepts and their implementation using an online platform called AgentBlock.net.

This tutorial addresses the use of stochastic gradients in simulation optimization, including methodology and algorithms, theoretical convergence analysis, and applications. Specific topics include stochastic gradient estimation techniques -- both direct unbiased and indirect finite-difference-based; stochastic approximation algorithms and their convergence rates; stochastic gradient descent (SGD) algorithms with momentum and reusing past samples via importance sampling; and a real-world application in geographical partitioning.

Optimization viewed broadly can aid in designing, analyzing, and operating complex systems, from strategic policy planning to last-mile distribution to optimal control of dynamical systems. The optimization model, including decision variables, objective functions, and constraints, requires performance metrics that are often evaluated via black-box simulations. We summarize algorithms that can address black-box noisy functions, mixed integer- and real-valued variables, and multiple objectives. Multiple models (e.g., Gaussian processes, neural networks, queueing networks) can be used in conjunction with a computationally expensive model (e.g., simulation) to predict performance and reduce overall computation. A key issue in solving an optimization model is to dynamically allocate computational effort to efficiently search for the global optimum. The dilemma of exploration vs exploitation vs estimation is evident in machine learning and global optimization. We discuss sampling distributions that provide insights into this balancing act, and how ideas in quantum optimization provide approaches to optimizing complex systems.

This tutorial introduces quantum approaches to Monte Carlo computation with applications in computational finance. We outline the basics of quantum computing using Grover's algorithm for unstructured search to build intuition. We then move slowly to amplitude estimation problems and applications to counting and Monte Carlo integration, again using Grover-type iterations. A hands-on Python/Qiskit implementation illustrates these concepts applied to finance. The paper concludes with a discussion on current challenges in scaling quantum simulation techniques.

Digital Twins, virtual models that accurately replicate the dynamic behavior of physical systems, are on the cusp of a revolutionary transformation not only for the design but also for real-time control of dynamic systems. In this tutorial, we define the key building blocks of Digital Twins and explore the key challenges they face in enabling real-time decision making. We discuss adaptive intelligence, an innovative reasoning framework that combines AI-driven approaches with simulation optimization techniques to enhance real-time decision making capabilities –hence, the value– of Digital Twins.

This tutorial introduces a new cuisine to the simulationist kitchen: plausible inference, the culinary art of output analysis that concocts statistical inferences about possibly unsimulated settings (e.g., solutions, parameters, decision variables) by blending and baking experiment design and problem structure. Tasty applications include screening out settings with undesirable or uninteresting performance or constructing confidence intervals and regions for a setting's measure(s) of performance. This tutorial synthesizes several disparate works on plausible inference into a cohesive, unifying framework with the aim of demonstrating how plausible inference recipes can satisfy a range of analyst appetites. Bon appetit!

The integration of Agent-Based Simulation (ABS) and Reinforcement Learning (RL) has emerged as a promising and effective approach for supporting decision-making in medical and hospital settings. This study proposes a novel framework that combines an Agent-Based Simulation with a Double Deep Q-Network (DDQN) Reinforcement Learning model to optimize task scheduling of healthcare professionals responsible for elderly patient care. Simulations were conducted over a 365-day period involving 250 patients, each managed by a healthcare coordinator who schedules appointments. Patients autonomously decide whether to attend appointments and adhere to medical recommendations. Results show the effectiveness of the RL model in minimizing health risks, with 84.8% of patients maintaining or improving their initial health risk levels, while only 15.2% experienced an increase.

Many countries plan to adopt hydrogen as a major energy carrier, which requires a robust infrastructure to meet rising demand. This paper presents a simulation model quantitatively analyzing the capacity of a potential hydrogen infrastructure in a coastal region of Northern Germany to supply a hydrogen hub in Bremen. The model covers ship-based imports of hydrogen, either as liquid hydrogen or ammonia, unloading at port terminals, conversion to gaseous hydrogen, pipeline transport to the hub, and end-use consumption. Various scenarios are simulated to quantitatively assess infrastructure needs under projected demand. Results show that ammonia-based imports offer greater supply reliability under low and medium demand, while liquid hydrogen performs better under high demand due to faster unloading times. Demand-driven supply policies generally outperform fixed-interval approaches by maintaining higher storage levels and aligning supply more closely with consumption patterns.

Agent-based social simulation (ABSS) has gained attention as a powerful method for analyzing complex social phenomena. However, the visualization of ABSS outputs is often difficult to interpret for users without expertise in ABSS modeling. This study analyzes how statistical literacy affects the comprehension of ABSS visualizations, based on cognitive processes defined in educational psychology. A web-based survey using five typical visualizations based on Schelling’s segregation model was conducted in Japan. The results showed a moderate positive correlation between statistical literacy and visualization comprehension, while some visualizations remained difficult to interpret even for participants with high literacy. Further machine learning analysis revealed that model performance varied by cognitive stage, and that basic and applied statistical skills had different impacts on comprehension across stages. These findings provide a foundation for designing visualizations tailored to user characteristics and offer insights for effective communication based on ABSS.

Advances in Artificial Intelligence have enabled more accurate and scalable modelling of complex social systems, which depend on realistic high-resolution population data. We introduce a novel methodology for generating hierarchical synthetic populations using differentiable programming, producing detailed demographic structures essential for simulation and analysis. Existing approaches struggle to model hierarchical population structures and optimise over discrete demographic attributes. Leveraging feed-forward neural networks and Gumbel-Softmax encoding, our approach transforms aggregated census and survey data into continuous, differentiable forms, enabling gradient-based optimisation to match target demographics with high fidelity. The framework captures multi-scale population structures, including household composition and socio-economic diversity, with verification via logical rules and validation against census cross tables. A UK case study shows our model closely replicates real-world distributions. This scalable approach provides simulation modellers and analysts with, high-fidelity synthetic populations as input for agent-based simulations of complex societal systems, enabling behaviour simulation, intervention evaluation, and demographic analysis.

Synthetic populations serve as the building blocks for predictive models in many domains, including transportation, epidemiology, and public policy. Therefore, using realistic synthetic populations is essential in these domains. Given the wide range of available techniques, determining which methods are most effective can be challenging. In this study, we investigate five synthetic population generation techniques in parallel to synthesize population data for various regions in North America. Our findings indicate that iterative proportional fitting (IPF) and conditional probabilities techniques perform best in different regions, geographic scales, and with increased attributes. Furthermore, IPF has lower implementation complexity, making it an ideal technique for various population synthesis tasks. We documented the evaluation process and shared our source code to enable further research on advancing the field of modeling and simulation.

Graph dynamical systems (GDSs) are widely used to model and simulate realistic multi-agent social dynamics, including societal unrest. This involves representing the multiagent system as a network and assigning functions to each vertex describing how they update their states based on the neighborhood states. However, in many contexts, social dynamics are triggered by external processes, which can affect the state transitions of agents. The classical GDS formalism does not incorporate such processes. We introduce the STP-GDS framework, that allows a GDS to be triggered by spatiotemporal background processes. We present a rigorous definition of the framework followed by formal analysis to estimate the size of the active neighborhood under two types of process distribution. The real-life applicability of the framework is further highlighted by an additional case study involving evacuation due to natural events, where we analyze collective agent behaviors under heterogeneous environmental and spatial settings.

This paper presents an open-source agent-based simulation model to study crowd-sourced last-mile food delivery. Within this context, we focus on a system that allows couriers with varying degrees of autonomy and cooperativeness to make decisions about accepting orders and strategically relocating. We model couriers as agents in an agent-based simulation model implemented in NetLogo. Our approach provides the necessary parameters to control and balance system performance in terms of courier productivity and delivery efficiency. Our simulation results show that moderate levels of autonomy and cooperation lead to improved performance, with significant gains in workload distribution and responsiveness to changing demand patterns. Our findings highlight the potential of self-organizing and decentralized strategies to improve scalability, adaptability, and fairness in platform-based food delivery logistics.

This paper applies an agent-based simulation model to examine the feasibility of battery electric trucks (BETs) in intermodal freight transportation, focusing on the Memphis hub network. Two infrastructure deployment stages, depot charging only and depot plus destination charging, are modeled and simulated using AnyLogic platform to study truck utilization patterns. Real-world manufacturing sites are chosen, and the trucks are routed along roadways using a Geographic Information System (GIS) map. Battery charge levels and charging infrastructure are modeled under both scenarios. Four electric truck models from various manufacturers including Tesla Semi, Nikola Tre, Volvo VNR, and Freightliner eCascadia are compared in terms of performance and utilization. Results showed that battery electric trucks are a feasible solution for intermodal trucking operations and transporting goods from manufacturers to destinations. This comparison also highlights effects of changing shifts and adding opportunity charging at destinations on truck utilization under different battery efficiencies and capacities.

We examine various sorting and selection methods for computing quantile and the conditional value-at-risk, two of the most commonly used risk measures in risk management scenarios. We study the situation where simulation data is already pre-generated, and perform timing experiments on calculating risk measures on the existing datasets. Through numerical experiments, approximate analyses, and existing theoretical results, we find that selection generally outperforms sorting, but which selection strategy runs fastest depends on several factors.

Nested simulation is a powerful tool for estimating widely-used risk measures, such as Value-at-Risk (VaR). While point estimation of VaR has been extensively studied in the literature, the topic of interval estimation remains comparatively underexplored. In this paper, we present a novel nested simulation procedure for constructing confidence intervals (CIs) for VaR with statistical guarantees. The proposed procedure begins by generating a set of outer scenarios, followed by a screening process that retains only a small subset of scenarios likely to result in significant portfolio losses. For each of these retained scenarios, inner samples are drawn, and the minimum and maximum means from these scenarios are used to construct the CI. Theoretical analysis confirms the asymptotic coverage probability of the resulting CI, ensuring its reliability. Numerical experiments validate the method, demonstrating its high effectiveness in practice.

We introduce the first fully automated fixed-sample-size procedure (FIRQUEST) for computing confidence intervals (CIs) for steady-state quantiles based on independent replications. The user provides a dataset from a number of independent replications of arbitrary size and specifies the required quantile and nominal coverage probability of the anticipated CI. The proposed method is based on the simulation analysis methods of batching, standardized time series (STS), and sectioning. Preliminary experimentation with the waiting-time process in an M/M/1 queueing system showed that FIRQUEST performed well by appropriately handling initialization effects and delivering CIs with estimated coverage probability close to the nominal level.

Latent Dirichlet Allocation (LDA) is a method for finding topics in text data. Evaluating an LDA model entails estimating the expected likelihood of held-out documents. This is commonly done through Monte Carlo simulation, which is prone to high relative variance. We propose an importance sampling estimator for this problem and characterize the theoretical asymptotic statistical efficiency it achieves in large documents. We illustrate the method in simulated data and in a dataset of news articles.

We develop exact importance sampling schemes for (linear) Hawkes processes to efficiently compute their tail probabilities. We show that the classical approach of exponential twisting, while conceptually simple to apply, leads to simulation estimators that are difficult to implement. These difficulties manifest in either a simulation bias or unreasonable computational costs. We mimic exponential twisting with an exponential martingale approach to achieve identical variance reduction guarantees but without the aforementioned challenges. Numerical tests compare the two, and present benchmarks against plain Monte Carlo.

The structure of financial networks plays a crucial role in managing financial risks, particularly in the assessment of systemic risk. However, the true structure of these networks is often difficult to observe directly. This makes it essential to develop methods for sampling possible network configurations based on partial information, such as node degree sequences. In this paper, we consider the problem of sampling bipartite graphs (e.g., bank-asset networks) under such partial information. We first derive exact bounds on the number of nodes that can be connected at each step, given a prescribed degree sequence. Building on these bounds, we then introduce a weighted-balanced random sampling algorithm for generating bipartite graphs that are consistent with the observed degrees, and illustrate how the algorithm works through an example. In addition, we demonstrate the effectiveness of the proposed algorithm through numerical experiments.

We propose an adversarial deep reinforcement learning (ADRL) algorithm for high-dimensional stochastic control problems. Inspired by the information relaxation duality, ADRL reformulates the control problem as a min-max optimization between policies and adversarial penalties, enforcing non-anticipativity while preserving optimality. Numerical experiments demonstrate ADRL’s superior performance to yield tight dual gaps. Our results highlight ADRL’s potential as a robust computational framework for high-dimensional stochastic control in simulation-based optimization contexts.

We consider estimating $p_x=P(X>x)$ in a data-driven manner or through simulation, when $x$ is large and when independent samples of $X$ are available. Naively, this involves generating $O(p_x^{-1})$ samples. Making distributional assumptions on $X$ reduces the sample complexity under commonly used distributional parameter estimators. It equals $O(\log p_x^{-1})$ for Gaussian distribution with unknown mean and known variance, and $O(\log^2 p_x^{-1})$ when the variance is also unknown and when the distribution is either exponential or Pareto. We also critically examine the more sophisticated assumption that the data belong to the domain of attraction of the Fréchet distribution allowing estimation methods from extreme value theory (EVT) . Our sobering conclusion based on sample complexity analysis and numerical experiments is that under these settings errors from estimation can be significant so that for probabilities as low as $10^{-6}$, naive methods may be preferable to those based on EVT.

Control variate (CV) is a powerful Monte Carlo variance reduction technique by injecting mean information about an auxiliary related variable to the simulation output. Partly due to this way of leveraging information, CV has been mostly studied in the context of mean estimation. In this paper, we study CV for general nonlinear quantities such as conditional value-at-risk and stochastic optimization. While we can extend the tools from mean estimation, a challenge in nonlinear generalizations is the proper calibration of the CV coefficient, which deviates from standard linear regression estimators and requires influence function or resampling. As a key contribution, we offer a general methodology that bypasses this challenge by harnessing a weighted representation of CV and interchanging weights between the empirical distribution and the nonlinear functional. We provide theoretical results in the form of central limit theorems to illustrate our performance gains and numerically demonstrate them with several experiments.

Consider estimating a known smooth function (such as a ratio) of unknown means. Our paper accomplishes this by first estimating each mean via randomized quasi-Monte Carlo and then evaluating the function at the estimated means. We prove that the resulting plug-in estimator obeys a central limit theorem by first establishing a joint central limit theorem for a triangular array of estimators of the vector of means and then employing the delta method.

Quasi-Regression (QR) is an inference method that approximates a function of interest (e.g., black-box model) for interpretation purposes by a linear combination of orthonormal basis functions of $L^2[0,1]^{d}$. The coefficients are integrals that do not have an analytical solution and therefore must be estimated, using Monte Carlo or Randomized Quasi-Monte Carlo (RQMC). The QR method can be time-consuming if the number of basis functions is large. If the function of interest is sparse, many of these basis functions are irrelevant and could thus be removed, but they need to be correctly identified first. We address this challenge by proposing new adaptive basis search methods based on the RQMC method that adaptively select important basis functions. These methods are shown to be much faster than previously proposed QR methods and are overall more efficient.

Computationally expensive combat simulations are often used to inform military decision-making and sensitivity analyses enable the quantification of the effect of military capabilities or tactics on combat mission effectiveness. The sensitivity analysis is performed using a meta-model approximating the simulation's input-output relationship and the output data that most combat meta-models are fitted to correspond to end-of-run mission effectiveness measures. However during execution, a simulation records a large array of temporal data. This paper seeks to examine whether functional combat meta-models fitted to this temporal data, and the subsequent sensitivity analysis, could provide a richer characterization of the effect of military capabilities or tactics. An approach from Functional Data Analysis will be used to illustrate the potential benefits on a case study involving a closed-loop, stochastic land combat simulation.

The performance of a CPU-only implementation of the restarted GMRES algorithm with direct randomized-SVD -based preconditioning has been analyzed. The method has been tested on a set of sparse and dense matrices exhibiting varying spectral properties and compared to the ILU(0) -based preconditioning. This comparison aims to assess the advantages and drawbacks of both approaches. The trade-off between iteration-to-solution and time-to-solution metrics is discussed, demonstrating that the proposed method achieves an improved convergence rate in terms of iterations. Additionally, the method’s competitiveness with respect to both metrics is discussed within the context of several relevant scenarios, particularly those where GMRES-based simulation techniques are applicable.

Gradient-based methods are well-suited for derivative-free optimization (DFO), where finite-difference (FD) estimates are commonly used as gradient surrogates. Traditional stochastic approximation methods, such as Kiefer-Wolfowitz (KW) and simultaneous perturbation stochastic approximation (SPSA), typically utilize only two samples per iteration, resulting in imprecise gradient estimates and necessitating diminishing step sizes for convergence. In this paper, we combine a batch-based FD estimate and an adaptive sampling strategy, developing an algorithm designed to enhance DFO in terms of both gradient estimation efficiency and sample efficiency. Furthermore, we establish the consistency of our proposed algorithm and demonstrate that, despite using a batch of samples per iteration, it achieves the same sample complexity as the KW and SPSA methods. Additionally, we propose a novel stochastic line search technique to adaptively tune the step size in practice. Finally, comprehensive numerical experiments confirm the superior empirical performance of the proposed algorithm.

Tackling simulation optimization problems with non-convex objective functions remains a fundamental challenge in operations research. In this paper, we propose a class of random search algorithms, called Regular Tree Search, which integrates adaptive sampling with recursive partitioning of the search space. The algorithm concentrates simulations on increasingly promising regions by iteratively refining a tree structure. A tree search strategy guides sampling decisions, while partitioning is triggered when the number of samples in a leaf node exceeds a threshold that depends on its depth. Furthermore, a specific tree search strategy, Upper Confidence Bounds applied to Trees (UCT), is employed in the Regular Tree Search. We prove global convergence under sub-Gaussian noise, based on assumptions involving the optimality gap, without requiring continuity of the objective function. Numerical experiments confirm that the algorithm reliably identifies the global optimum and provides accurate estimates of its objective value.

This paper addresses the challenge of reliably selecting high-performing solutions in simulation optimization when model parameters are uncertain. We infer the set of solutions whose probabilities of performing better than a user-defined threshold are above a confidence level given the uncertainty about the parameters. We show that this problem can be formulated as a nested superlevel set estimation problem and propose a sequential sampling framework that models the simulation output mean as a Gaussian process (GP) defined on the solution and parameter spaces. Based on the GP model, we introduce a set estimator and an acquisition function that evaluates the expected number of solutions whose set classifications change should each solution-parameter pair be sampled. We also provide approximation schemes to make the acquisition function computation more efficient. Based on these, we propose a sequential sampling algorithm that effectively reduces the set estimation error and empirically demonstrate its performance.

Missing data presents a persistent challenge in machine learning. Conventional approaches often rely on data imputation followed by standard learning procedures, typically overlooking the uncertainty introduced by the imputation process. This paper introduces Imputation-based Distributionally Robust Logistic Regression (I-DRLR)—a novel framework that integrates data imputation with class-conditional Distributionally Robust Optimization (DRO) under the Wasserstein distance. I-DRLR explicitly models distributional ambiguity in the imputed data and seeks to minimize the worst-case logistic loss over the resulting uncertainty set. We derive a convex reformulation to enable tractable optimization and evaluate the method on the Breast Cancer and Heart Disease datasets from the UCI Repository. Experimental results demonstrate consistent improvements for out-of-sample performance in both prediction accuracy and ROC-AUC, outperforming traditional methods that treat imputed data as fully reliable.

This study explores strategies for robust optimization of queueing performance in the presence of input model uncertainty. Ambiguity sets for Distributionally Robust Optimization (DRO) based on Wasserstein distance is preferred for general DRO settings where the computation of performance given the distribution form is straightforward. For complex queueing systems, distributions with large Wasserstein distance (from the nominal distributions) do not necessarily provide extreme objective values. Thus, the calculation of performance extremes must be done via an inner level of maximization, making DRO a compute-intensive activity. We explore approximations for queue waiting time in a number of settings and show how they can provide low-cost guidance on extreme objective values, allowing for more rapid DRO. Approximations are provided for single- and multi-server queues and queueing networks, each illustrated with an example. We also show in settings with small number of solution alternatives that these approximations lead to robust solutions.

Expected improvement (EI) is a common ranking and selection (R&S) method for selecting the optimal system design from a finite set of alternatives. Ryzhov (2016) observed that, under normal sampling distributions with known variances, the limiting budget allocation achieved by EI was closely related to the theoretical optimum. However, when the variances are unknown, the behavior of EI was quite different, giving rise to the question of whether the optimal allocation in this setting was totally distinct from the known-variance case. This research solves that problem with a new analysis that can distinguish between known and unknown variance, unlike previously existing theoretical frameworks. We derive a new optimal budget allocation for this setting, and confirm that the limiting behavior of EI has a similar relationship to this allocation as in the known-variance case.

This study explores the integration of machine learning (ML) clustering techniques into a simulation-optimization framework aimed at enhancing the efficiency of airport security checkpoints. Simulation-optimization is particularly suited for addressing problems characterized by evolving data uncertainties, necessitating critical system decisions before the complete data stream is observed. This scenario is prevalent in airport security, where passenger arrival times are unpredictable, and resource allocation must be planned in advance. Despite its suitability, simulation-optimization is computationally intensive, limiting its practicality for real-time decision-making. This research hypothesizes that incorporating ML clustering techniques into the simulation-optimization framework can significantly reduce computational time. A comprehensive computational study is conducted to evaluate the performance of various ML clustering techniques, identifying the OPTICS method as the best found approach. By incorporating ML clustering methods, specifically the OPTICS technique, the framework significantly reduces computational time while maintaining high-quality solutions for resource allocation.

Aircraft manufacturing presents significant challenges for logistics departments due to the complexity of processes and technology, as well as the high variety of parts that must be handled. To support the development and optimization of these complex logistics processes in the aviation industry, simulation is commonly employed. However, existing simulation models are typically tailored to specific use cases. Reusing or adapting these models for other aircraft-specific applications often requires substantial implementation and validation efforts. As a result, there is a need for flexible and easily adaptable simulation models. This work aims to address this challenge by developing a modular library for logistics processes in aircraft manufacturing. The outcome of this work highlights the simplifications introduced by the developed library and its application in a real aviation warehouse.

This paper introduces a methodology for simulating flight dynamics utilizing the Tool Command Language (Tcl). Tcl, created by John Ousterhout, was conceived as an embeddable scripting language for an experimental Computer Aided Design (CAD) system. Tcl, a mature and maturing language recognized for its simplicity, versatility, and extensibility, is a compelling contender for the integration of flight dynamics functionalities. The work presents an extension method utilizing Tcl's adaptability for a novel type of flight simulation programming. Initial test findings demonstrate performance appropriate for the creation of human-in-the-loop real-time flight simulations. The possibility for efficient and precise modeling of future complicated distributed simulation elements is discussed, and recommendations regarding subsequent development priorities are drawn.

Discrete Event Simulation (DES) can be a powerful tool to help companies make better decisions. However, just one incorrect application of simulation can leave a bad taste with project sponsors and can derail future simulation endeavors. What are appropriate uses for simulation? What type of model do you need and how will that drive the data that should be collected? This presentation will help you get your project started on the right foot by giving you tips on what things to consider before kicking it off.

The integration of Computer-Aided Design (CAD) models into discrete event simulation software is a critical requirement for many simulation projects, particularly those involving the movement of people or vehicles where spatial accuracy directly impacts study outcomes. While importing CAD files and configuring simulation elements is essential for system accuracy, this process is typically time-consuming, prone to errors, and involves substantial repetitive tasks. Although previous attempts have been made to automate this workflow, the wide variety of CAD formats and lack of standardization pose significant challenges. This paper presents novel approaches for automating CAD import processes, specifically focusing on 2D drawings using the ezdxf Python library (ezdxf) and 3D models using Revit Python Shell (revitpythonshell). Our methods demonstrate potential time savings and error reduction in simulation model development while maintaining spatial accuracy.

One of North America’s largest convenience store chains launched a strategic initiative to expand fresh food offerings across all locations, requiring significant changes to store layouts, equipment, staffing, and operations. To support data-driven decision-making for these remodels, the client partnered with MOSIMTEC to develop a virtual twin of its most common store layout using AnyLogic simulation software. The simulation model, integrated with Excel and PowerBI, enabled stakeholders to evaluate store configurations, labor needs, equipment impacts, and throughput performance—ultimately guiding smarter investments in their Fresh Food program.

Dynamic work movement is a transformative approach that allows tasks not fully completed at their designated position to transition seamlessly to subsequent positions within the workflow. This method ensures that delays are minimized and the entire process remains fluid and dynamic. It is particularly useful for tasks not entirely bound by their positional location, providing remarkable flexibility and adaptability in the workflow process. Facilitating the movement of jobs that are behind schedule to follow-on positions, dynamic work movement ensures that the production process remains on schedule, thereby enhancing overall efficiency and productivity. The implementation of dynamic work movement significantly improves the flexibility and adaptability of the production system. The streamlined process ensures continuous progress and minimizes delays, ultimately leading to a more efficient and effective production system. This approach also enhances the ability to handle variability and disruptions, ensuring that the production system can adapt to changing conditions and requirements.

This case study examines the implementation of a Simio-based digital twin scheduling model for a major Australian snack food manufacturer. The company faced significant scheduling challenges due to complex production constraints, including shared assets, intricate product routing, and the transition between multiple manufacturing sites. The previous Excel-based scheduling system could not adequately handle these complexities, resulting in suboptimal resource utilization and difficulty in scenario planning. The Simio model incorporated flow-based modeling techniques to represent fryer outputs and bagger draws, enabling dynamic scheduling that respected all production constraints while optimizing resource utilization. The implementation resulted in improved planning efficiency, better alignment between departments, enhanced capacity understanding, and support for the successful commissioning of a new manufacturing facility. This case demonstrates how simulation-based scheduling can address complex manufacturing environments with multiple interdependent constraints.

The shipbuilding industry faces challenges in simulation-based production forecasting due to custom design requirements, design-to-production variability, labor-intensive processes, and complex scheduling. Increasing labor shortages and cost pressures highlight the need for a data-driven operational framework. A discrete-event simulation model tailored for ship block assembly planning is presented, defining standardized work units and deployment criteria to forecast labor and equipment workload and schedule delays. The simulator significantly shortens planning and load analysis time, improves delay forecasting accuracy, and visualizes load distribution and bottleneck patterns through a user-friendly interface. These capabilities enhance management efficiency and decision-making confidence in practical shipyard operations.

This lively and candid panel, Challenging Projects: Learn from My Mistakes, invites seasoned industry practitioners to share their most painful modeling missteps and project failures, along with the hard-earned lessons that came from them. Rather than focusing on polished success stories, panelists will reflect on what went wrong, why it happened, and what they would do differently with the benefit of hindsight. The goal is to broaden the conversation beyond individual case studies and extract practical wisdom from the real-world experiences of simulation professionals. New modelers will gain valuable insight into common pitfalls, unspoken challenges, and the importance of resilience and reflection in building a successful career.

Traditional automotive painting, such as spray painting with oven drying, faces scalability challenges, high energy use, and significant environmental impacts. Emerging alternatives include ink‑jet printing with UV curing and painted‑film vacuum forming. To fully capitalize on these advanced technologies, feasibility assessments through advanced simulation and optimization techniques are essential at the design stage of manufacturing systems. This study focuses on two processes: (1) the ink-jet printing & UV light curing-based painting process and (2) the painted film vacuum forming painting process. We aim to identify potential operational issues and propose corresponding solutions to ultimately achieve desired throughputs. It is expected that the considerations presented regarding production operations for adopting advanced painting technologies will serve as a useful reference for practitioners in the automotive painting area.

This paper presents a hybrid framework that combines discrete-event simulation (DES) with neural networks to forecast Work-In-Progress (WIP) in semiconductor fabs. The model integrates three learned components: a dispatching model, an inter-start time predictor, and a processing time estimator. These models drive a lightweight simulation engine that accurately predicts WIP across various aggregation levels.

This case study examines how LMAC Group utilized Simio simulation software to support a New Zealand metal fabrication manufacturer in scaling production from 600 to 26,000 units while maintaining onshore manufacturing and reducing unit costs. The simulation model enabled analysis of current production capacity, identification of constraints, and evaluation of mitigation strategies including shift pattern modifications and automation options. The approach demonstrated how simulation could inform capital investment decisions by providing data-driven insights before physical implementation, resulting in optimized worker allocation, improved machine utilization, and a comprehensive future state factory design.

Stochastic discrete event simulation is a vital tool across industries. However, the high dimensionality and complexity of real-world systems make it challenging to develop simulations that accurately model and predict business metrics for users when faced with inaccurate input and model fidelity limitations. Addressing this challenge is critical for improving the effectiveness of industrial simulations. In this work, we focus on simulation output uncertainty, a crucial summary statistic for assessing business risks. We introduce a novel framework called INput-aware Quantification of Uncertainty for Interpretation, Risk, and Experimentation (INQUIRE). At the heart of INQUIRE, we develop a residual-based uncertainty prediction model driven by key input parameters. Then we incorporate a skewness-detection procedure for quantile estimation that provides risk assessment. To analyze how input parameters evolution influences simulation output uncertainty, we introduce a Shapley-value-based interpretation method. Additionally, our framework enables more efficient simulation-driven experimentation, enhancing strategic decision-making by providing deeper insights.

Enterprise-level simulation platforms model complex systems with thousands of interacting components, enabling organizations to test hypotheses and optimize operations in a virtual environment. Among these, supply chain simulations play a crucial role in planning and optimizing complex logistics operations. As these simulations grow more sophisticated, robust methods are needed to explain their outputs and identify key drivers of change. In this work, we introduce a novel causal attribution framework based on the Shapley value, a game-theoretic approach for quantifying the contribution of individual input features to simulation outputs. By integrating Shapley values with explainable Gaussian process models, we effectively decompose simulation outputs into individual input effects, improving interpretability and computational efficiency. We demonstrate our framework using both synthetic and real-world supply chain data, illustrating how our method rapidly identifies the root causes of anomalies in simulation outputs.

We propose a new method to efficiently simulate customer-averaged service metrics in nonstationary queueing systems with time-varying arrivals and staffing. Key delay-based metrics include the fraction of customers waiting no more than x minutes, average waiting time, and abandonment rate over a finite horizon. Traditional discrete-event simulation (DES) tracks individual customers, leading to complex implementation and high variance due to numerous random events. To address this, we introduce a queue-based estimator that conditions on the time-dependent queue-length process, significantly reducing variance and simplifying the implementation of the simulation. We further enhance this estimator using many-server heavy-traffic approximations to capture queue dynamics more accurately. Numerical results show our method is much more computationally efficient than standard DES, making it highly scalable for large systems.

Verification and Validation (V&V) are foundational to building trust in simulation models, yet their application in practice frequently involves ambiguity and difficulty. This case study explores how simulation practitioners navigate the complexities of V&V across different industries. Drawing on academic frameworks, practitioner interviews and industry case studies, we examine the disconnect between textbook theory and real-world implementation.

Synthetic populations with spatially connected individuals are useful for modeling infectious diseases, particularly when assessing the impact of interventions on geographic and demographic subgroups. However, open-source tools that incorporate geospatial data are limited. We present a method to generate synthetic contact networks for any US region using publicly available data. GeoPops is an open-source package that 1) synthesizes a set of agents approximating US Census distributions, 2) assigns agents to homes, schools, and workplaces, and 3) connects agents within these locations. We compare GeoPops to two other synthesis tools that generate comparable network structures by creating Maryland populations with each package and assessing accuracy against US Census data. We simulate COVID-19 transmission on each population and compare simulations to observed data for the first wave of the pandemic. Our study highlights the utility of spatially connected synthetic populations and builds the capacity of modelers to better inform epidemic decision making.

This paper examines Lockheed Martin’s implementation of a Training Enterprise Digital Twin system that revolutionizes military training operations management. Using Simio, Lockheed Martin created a comprehensive, data-driven digital model of a training enterprise to optimize resource allocation, forecast program performance, and support strategic decision-making. The simulation framework incorporates student progression through training pipelines alongside asset availability and maintenance requirements. Results demonstrate significant improvements in training efficiency via optimized training times, improved resource utilization, and substantial cost avoidance through optimized asset procurement. This case study illustrates how simulation technology enables defense contractors to meet performance-based contract requirements while maximizing return on investment in training operations.

Lucid Motors has begun a proof of concept which aims to formulate a motor thermal estimator using a modern time series transformer and integrate it into our current physics-based model, creating a “data digital twin” of the motor thermal system. Outputs include component-level heat and power loss estimates over time, which are important for tuning motor performance. Our goal with the data digital twin is to improve latency and increase the time step frequency for thermal estimation, which are currently limited by computational load and in turn limit the optimization and accuracy of control algorithms designed to react to the temperature gradients within the drive unit.

Many countries and companies are pursuing the goal of climate neutrality, which is increasing the importance of green hydrogen. However, due to high electricity and capital costs—particularly for electrolyzers—the production of green hydrogen remains expensive. As part of the StaR research project, a novel stack for alkaline electrolyzers was developed, along with a scalable production concept. To reduce investment risk, a semi-automated pilot production facility was established at the Dortmund site. The production processes and workforce allocation were modeled using a discrete-event simulation and analyzed with regard to productivity and resource assignment. A generic control logic enables flexible evaluation of various scenarios for work plan optimization. Real-world process data are continuously collected and integrated into the simulation model for ongoing updates.

Westinghouse Electric Company partnered with MOSIMTEC to implement Simio-based digital twins, replacing fragmented, Excel-driven planning across five nuclear fuel production sites. The solution integrates data from SAP, IMS, and Excel via ETL processes to enable dynamic, capacity-constrained scheduling and real-time scenario analysis. Developed through a structured methodology—including functional specification, phased modeling, validation, and integration, the model supports multi-site planning, throughput analysis, material flow visualization, and resource utilization reporting. Planning cycles were reduced from weeks to hours, improving responsiveness to customer requests, outages, and shifting priorities. The project revealed data inconsistencies across systems, prompting quality improvements. Westinghouse planners now independently run and modify plans without external support, enabling faster decisions and improved cross-site coordination. The digital twins have established a scalable, data-driven framework that supports ongoing optimization and future expansion into other business units. This case highlights how simulation can enable operational agility in complex, high-regulation manufacturing environments.

Online retailer supply chain management involves decisions about how much and when to buy inventory, where to inbound new inventory, how to transfer inventory between warehouses, and how to fulfill customer orders. These choices must adhere to capacity constraints, such as labor plans, while minimizing impacts on customer service and profitability. This paper presents a dual decomposition framework and simulation-based cost search approach that maintains high customer service levels while better aligning buying purchase orders with inbound node arrival capacity plans. The methodology is evaluated through end-to-end discrete event simulations, which quantify the impact of candidate capacity costs on buying system behavior, as well as the downstream effects on other systems and ultimately on customer service metrics. Results demonstrate this approach can improve capacity adherence by over 60% across a five-week horizon, with less than a 4% degradation in a metric measuring the proximity of inventory to customers.

Simulation plays a central role in the strategic planning and operational evaluation of supply chain networks. Within these networks, order fulfillment traditionally requires solving computationally expensive optimization problems in real-time across multiple constraints. For forward-looking simulations evaluating millions of orders, such optimization becomes prohibitively expensive. We develop a neural network-based emulator that approximates optimal fulfillment decisions while maintaining millisecond-level inference speed. Operating at ZIP-code level resolution and incorporating shipping speed constraints, our model handles exponential decision spaces and non-stationary patterns. Empirical results demonstrate 56.75% order-level accuracy, a 20 percentage point improvement over benchmarks. Through novel regularization balancing order-level and network-level efficiency, we achieve 47.13% node-level accuracy while maintaining 50.31% order-level accuracy. Our model captures intricate patterns in historical fulfillment data, enabling efficient forward-looking simulation for strategic planning.

For large-scale retail businesses such as Amazon and Walmart, simulation is critical for forecasting inventory, planning, and decision-making. Traditional digital-twin simulators, which are powered by complex optimization algorithms, are high-fidelity but can be computationally expensive and difficult to experiment with. We propose a hybrid simulator called SEmulate that integrates machine learning and emulation into large-scale simulation systems that follows the causal relationship between system components. SEmulate uses machine learning to emulate complex high-dimensional system components, while using physics-based modeling or heuristics for other simpler components. We apply SEmulate on supply chain simulation, where we demonstrate that SEmulate reduces runtime by orders of magnitude compared to the traditional digital-twin based simulator, while maintaining competitive accuracy and interpretability. The enhanced simulator enables efficient supply chain simulation, rapid experimentation, faster and improved decision-making process.

Amazon's sort centers represent highly complex systems where traditional monolithic simulation approaches fall short due to computational limitations, extended time-to-value, and challenging model maintenance. This case study presents a federated simulation framework that decomposes the sort center model into four key federates: inbound operations, main sortation and outbound operations, non-conveyable processing, and non-compliant item handling. Each federate operates independently while maintaining synchronization through a conservative time-stepping algorithm and structured message-passing protocol. Our implementation demonstrates significant advantages, including a 40% reduction in model development time, 15% improvement in model execution time, and improved scenario testing capabilities. This approach enables early optimization of sort center design, identification of cross-process bottlenecks, and better first-time launch quality. This real-world application showcases the practicality and benefits of federated simulation in modern logistics, offering valuable insights for practitioners modeling complex industrial systems.

This case study examines how Penske Truck Leasing utilized Simio simulation software to address capacity planning challenges associated with fleet growth. Facing the addition of 500 vehicles over five years to an already space-constrained facility, Penske’s Operational Excellence team developed a comprehensive simulation model to identify capacity constraints and evaluate potential solutions. The model analyzed multiple capacity dimensions including parking space, service bays, technician staffing, and support resources. Through simulation, Penske identified specific capacity ceilings, determined optimal timing for implementing various solutions, and provided facility managers with a data-driven roadmap for supporting growth while maintaining operational efficiency. This approach enabled Penske to make informed decisions about resource allocation, facility modifications, and staffing adjustments without the risks associated with real-world implementation.

In the shipbuilding industry, ships are constructed from large, prefabricated blocks on fixed assembly workspace within a factory. The efficiency of the entire production schedule is highly dependent on how these blocks are arranged, as each has a unique geometry and a specific construction timeline. This study addresses this complex challenge by developing an optimization algorithm for block arrangement. The proposed method utilizes the No-Fit Polygon (NFP) to handle complex geometric constraints and employs Simulated Annealing (SA) to find a near-optimal arrangement that enhances overall workflow efficiency.

To lead in delivering safer, premium-quality food ingredients, one of our plants is undergoing a strategic transformation. This initiative reflects a forward-thinking commitment to innovation and excellence, not just regulatory alignment. By transitioning to a new class of high-quality products, we are enhancing safety, expanding our portfolio, and positioning ourselves to better serve evolving market needs. This shift introduces operational complexities - frequent product changeovers, tighter inventory control, and outbound capacity constraints. To address these, we developed a robust simulation model, offering comprehensive view of plant operations. The model enables us to test new configurations, quantify cost impacts, and proactively identify supply risks. Beyond immediate improvements, the simulation supports smarter infrastructure investments by evaluating storage expansion and modeling demand growth scenarios. This data-driven approach strengthens long-term flexibility and resilience, ensuring our operations remain agile and aligned with future market dynamics.

How do we know if the results of a simulation analysis are useful, accurate, reliable and credible? This is an important and fundamental question. If a simulation produces invalid results, yet the client accepts and acts upon them, then they’re heading into a world of pain. Just as bad is the situation in which the results are valid but are distrusted by the client and so are disregarded. In both cases, the outcomes can be catastrophic for the client. If the model is invalid, and the client rightly rejects it, then we’ve wasted our time and dented our credibility. The usefulness, accuracy and reliability of a simulation model—and the client’s trust in its results—are key factors for a successful outcome; this presentation will demonstrate how project objectives should drive the model’s content, its verification & validation, and experimentation.

This study uses discrete-event simulation in MineTwin to evaluate the feasibility of the strategic restart plan for South Crofty, a historic underground tin mine in Cornwall, UK. The original plan, developed with a static scheduling tool, did not capture equipment interactions, queuing delays, or operational variability. To overcome these limitations, we built a simulation model incorporating detailed mine layouts, production schedules, equipment specifications, and equipment coordination rules. The simulation revealed several constraints not identified in the static plan, including the need for additional loaders and insufficient surface haulage capacity. The simulation also enabled level-specific equipment allocation, which proved critical in pre-production phases with limited level connectivity. Overall, simulation complemented static planning by revealing dynamic bottlenecks and improving confidence in the operational feasibility of the restart plan.

This case study examines the application of discrete-event simulation to optimize operations at Emory Healthcare’s Dunwoody Family Medicine clinic, a resident-driven healthcare facility. The clinic faced challenges with patient wait times and complex resident-preceptor interactions that impacted patient flow. Using Simio simulation software, researchers developed a digital twin model of clinic operations, incorporating data from electronic health records and time studies. The validated model identified key bottlenecks and tested multiple improvement scenarios. Results showed that implementing a first-come-first-served preceptor queue system could reduce preceptor waiting time by 31%, while strategic resident pod assignments could decrease travel time by 60%. This project demonstrates how simulation modeling can provide healthcare facilities with data-driven insights to improve operational efficiency while maintaining educational quality in teaching environments.

In complex aluminum manufacturing systems, optimizing interdependencies is critical to avoiding unnecessary capital investment and costly inefficiencies. At Novelis remelt and recycling plants, crane operations are central to metal movement and production flow, creating challenges when operational capacity is limited. To address this, the Novelis simulation team developed a 3D flow model in AnyLogic that mirrors crane operations, incorporating factors like processing times, schedules, and downtime. The innovative simulation combined real-world data with dynamic modeling to evaluate crane capacity under varying conditions. Unlike static methods, the cloud-deployed model featured an intuitive interface, allowing plant teams to test scenarios and identify capacity thresholds. Initial results showed the existing crane could meet increased production goals through reliability improvements and staffing changes, avoiding multimillion-dollar capital expenditure. The extended abstract will outline the simulation methodology, model architecture, interface features, and outcomes that enabled the plant’s successful adoption of the tool.

This case study examines how Seegrid, a leading manufacturer of autonomous mobile robot (AMR) solutions, partnered with Simio to develop a comprehensive discrete event simulation template model that reliably represents Seegrid AMR solutions, enabling better visualization of vehicle interactions, congestion reduction, and optimal fleet sizing. Simio’s template-based approach enables Seegrid’s large team of skilled Application Engineers to quickly and accurately build models of customer automated workflows. The collaboration demonstrates how simulation can transform solution design for modern industrial autonomous material handling, providing both immediate operational insights and a foundation for future innovation.

This case study presents a discrete event simulation model of a robotic mobile fulfillment center (RMFC) built in Simio. The model evaluates how different inventory slotting strategies affect operational performance at three facility scales: a small, medium, and large system. Inspired by a shoe fulfillment center, the model features standardized shelf sizes and a large number of SKUs due to variations in shoe models, sizes, and colors. A synthetic dataset was created to reflect realistic order volumes and SKU diversity without relying on proprietary data. Key performance metrics include order cycle time, pile-on, and AGV travel distance. The goal is to demonstrate how simulation enables robust evaluation of storage strategies in variable-rich RMFC environments. This work provides a foundation for comparing slotting methods and exploring how their effectiveness changes as facility size scales.

This panel, Simulation in Industry: Past, Present, & Future, takes a wide-angle view of the barriers that continue to limit the use of simulation for major capital expenditure (Capex) decisions, with a focus on sectors like manufacturing, material handling and automation. Rather than spotlighting individual case studies, the discussion will explore systemic and cultural obstacles that slow adoption e.g. such as organizational inertia, messaging gaps, and misaligned expectations between technical teams and business leaders. In addition to diagnosing current challenges, panelists will offer reflections on how industrial simulation has evolved, where it stands today, and what changes are needed to ensure it plays a more central role in decision-making by 2050. This session invites the community to think critically about how we frame the value of simulation and how we must evolve to meet the future.

Cell therapy manufacturing presents unique challenges due to its high complexity, stringent regulatory requirements, and patient-specific variability. To enhance production efficiency and ensure robust scalability, simulation modeling emerges as an effective strategy to evaluate and refine manufacturing processes without disrupting ongoing operations. This study introduces a comprehensive framework that integrates discrete-event simulation (DES) to model critical cell therapy manufacturing stages—including inbound receipt and verification, QC sampling, material and kit-building, media fill process, manufacturing, cryopreservation, and outbound logistics. The framework not only captures resource constraints, batch scheduling, equipment utilization, and potential failure modes but also identifies the root causes of process bottlenecks. Further, it provides a roadmap for scenario design and scale-up analysis, enabling systematic sensitivity evaluations to quantify impacts on throughput, cost, and lead time. The findings underscore the framework’s potential to support capacity planning, workforce allocation, and quality risk management, thereby accelerating commercial readiness and improving patient access.

This case study explores the use of Discrete Event Simulation (DES) to optimize resource planning and patient flow management in a high-volume oncology infusion center anticipating a five-year surge in demand. Conducted at HonorHealth Cancer Center, the simulation model, built with MedModel software (BigBear.ai) incorporated empirical data and subject matter expertise to evaluate several strategic initiatives, including extended hours of operation, inter-clinic chair sharing, optimized patient and pharmacy hood scheduling, and front office staffing adjustments. The five-year planning scenarios revealed that, without intervention, the center would experience severe capacity limitations. However, the integration of targeted improvements significantly enhanced chair availability, reduced patient queue times, and increased overall operational efficiency. This study highlights the transformative role of DES as a decision-support tool, underscoring the importance of a DES consultant to join facility planning and design teams.

In simulation projects, model complexity often competes with limited resources, tight timelines, and stakeholder expectations. Effective scoping is essential to ensure efforts are focused on the right questions and the appropriate level of detail. A well-defined scope guides decisions on system boundaries, performance metrics, modeling approaches, and abstraction levels thereby supporting sound analysis while protecting the modeler from scope creep and misaligned expectations. Within this framework, the ability to make thoughtful simplifying assumptions becomes a critical modeling skill. When carefully applied, these assumptions preserve model validity and decision-making utility while improving development speed, stakeholder communication, and model ownership. This presentation shares practical strategies for scoping simulation projects and applying assumptions effectively, illustrated through real-world examples.

Enhancing operational efficiency while ensuring worker safety is a key objective in industrial system design. It is particularly important in environments where human workers and forklifts share the same physical space. Rolling stock factories often include warehouse areas that manage both large and small parts, relying on a combination of workers and forklifts to perform material transport. These warehouses play a critical role in supporting production by executing kitting operations, which represent the initial step in the assembly process and demand significant coordination and time management. This study employs discrete-event simulation (DES) to evaluate the efficiency and safety of various warehouse layout and process configurations prior to implementation. A lightweight industrial simulator, built on the Rapid Modeling Architecture (RMA), enabled rapid iterations of simulation-based analyses and supported effective decision-making before physical changes on the factory floor.

The final test operation in semiconductor production, which verifies the electrical functionality of packaged chips, demands highly responsive logistics and equipment control strategies to meet delivery, ensure quality, and accommodate urgent requests. In complex production environments involving diverse products and many different process constraints, optimizing both lot merging efficiency and equipment utilization becomes more and more critical. This paper presents an integrated framework that combines digital twin simulation with deep reinforcement learning, a Rainbow DQN agent, to enhance success rates of lot merging and improve throughput. While preserving the existing MES-driven dispatch logic, the framework introduces a learning-based policy that leverages equipment characteristics—particularly the constraint that higher-parallelism tools have narrower merging time windows. The agent learns to dynamically adjust lot-to-tool assignments, reducing resource waste and idle time, while supporting more efficient and predictable production scheduling.

As automotive manufacturing systems grow in complexity, the demand for efficient and dependable PLC programming continues to rise. Conventional manual programming is reliant on individual expertise so that leading to many points of human error. This study introduces a framework that automates PLC code generation using flowchart-based logic and validates the output through simulation. The approach translates flowcharts into IEC 61131-3 compliant ladder logic via domain-specific language and verifies functionality using 3D simulation environments integrated with OPC UA protocols. This method improves consistency, shortens development cycle of control logic, and reduces programming errors. Its effectiveness and scalability are demonstrated through a real-world application in an electric vehicle body shop process.

A manufacturer was considering a temporary operation in a constrained space. Simple calculations showed a possible bottleneck at the dock doors. An alternative was to build a connector to an adjacent warehouse, at significant capital expense, to improve flexibility and throughput. A discrete-event simulation was built to evaluate the benefit of investment in the connector. The model demonstrated the material handling constraint and that the connector had payback potential.

Demand for an established paint facility is expected to change significantly due to legacy product phasing out overtime and the introduction of a new product. The new product is physically larger, changing the rate parts can flow through the paint facility. Two primary questions were asked: 1) Can existing paint facilities meet future demand, and if not, when will demand exceed capacity? 2) If new paint booth technology were deployed, how much paint facility will be required? Two discrete event simulation models were developed to answer each question. The current state model played a primary role in identifying that demand would exceed capacity before a new system could be installed and then quantified the impact of implementing a provisional booth. The future state model quantified the resources required in various demand profiles and equipment configurations to ensure proper throughput and process lead-times.

Discrete Event Simulation (DES) frameworks have become essential tools in Amazon’s logistics network for analyzing and optimizing warehouse operations. This paper presents a modular DES framework developed using FlexSim to model operations across multiple Amazon IXD facilities. Validated against historical data, the model accurately replicates operational dynamics and supports applications span from current building optimization, particularly in cross-belt sorter recirculation parameters, to testing future automation concepts. Key challenges including human operational behavior and unexpected operational events are discussed. The study concludes by highlighting future directions in emulation and network model connectivity to enhance simulation fidelity and enable real-time data exchange across the logistics network.

This presentation aims to provide guidance on navigating the technology stack of digital transformation — with a focus on the complex, multifaceted relationship between Simulation and AI. We’ll discuss three key topics: 1) Why hasn’t simulation, as a training environment for machine learning, taken off at scale? 2) Can AI (and LLMs in particular) replace simulation? 3) How does AI affect the simulation model lifecycle, and how can it assist the modeler? We will also touch on the evolving requirements today’s industry places on simulation technologies.

This study proposes a simulation-based optimization framework and its application to vehicle fleet sizing in automated material handling systems (AMHSs). The framework comprises two stages: simulation calibration and optimization. The simulation calibration adjusts the simulation input parameters using field observations to obtain a more reliable simulation model. The simulation optimization determines the best number of vehicles to minimize flowtime under performance constraints based on the calibrated simulation. We validated the framework through application to a display fabrication plant in South Korea. The framework effectively modulates the fleet size in response to material flow and exhibits stable performance in the real-world AMHS.

This case study presents how Dijitalis Consulting utilized simulation modeling to optimize Automated Guided Vehicle (AGV) investments for a major electronics manufacturing facility. The client, a TV manufacturing company with 15 assembly lines, was initiating significant investments including a new AGV fleet to replace their outdated system of 132 vehicles. Using Simio simulation software, Dijitalis created a comprehensive digital model of the facility’s material handling operations, analyzing traffic patterns, potential bottlenecks, and resource utilization. The simulation revealed that only 95 AGVs were required instead of the initially proposed 132, resulting in capital expenditure savings exceeding $1.5 million. Beyond cost savings, the simulation model became a continuous improvement tool for testing layout modifications, process changes, and production schedule feasibility. This case demonstrates how simulation-based decision making can prevent overinvestment while ensuring operational requirements are met.

Ocado Technology developed On-Grid Robotic Pick (OGRP) - a robotic picking solution for grocery fulfillment centers - with coordinated innovation across hardware, software, and system design. To support this development, we built a low-fidelity simulation model early and evolved it alongside the product. As OGRP matured, assumptions from each area were progressively integrated into the model, while simulation insight shaped design decisions - forming a “product digital twin”.

We describe how this approach of co-evolving the simulation supported exploration of a large and uncertain design space - aligning development streams, guiding rapid iteration, and enabling early development - ultimately driving faster delivery with lower risk.

OGRP is now deployed around the world, picking over a million items per week and continuing to scale. This case study illustrates how evolving, system-level simulations can drive coordination in complex, fast-changing engineering projects.

The simulation industry has its own terms that are often used inconsistently between professionals. For example, much time is spent trying to define what makes a digital twin different from a simulation model. Inconsistent use of simulation terminology often leads to project dysfunctions, such as scope creep, misaligned expectations and solutions that fail to meet business objectives. In this presentation, we will present a wide range of terms representing the most common modeling techniques. The focus is to provide simulation professionals with varying definitions of the same term, so they can be prepared to ask the right questions to ensure project success. While this presentation focuses on clarifying framework terminology more than ideal framework selection, attendees will gain a better understanding of various frameworks being used in simulation modeling.

This extended abstract examines the architectural considerations for data entry in sustainment Modeling and Simulations (M&S). It highlights the necessity of having a flexible yet robust data architecture that can accommodate various data types, including early design engineering estimates, test data, and post-deployment maintenance data, which often come in different formats. By establishing a data structure that facilitates efficient extraction, transformation, and loading from multiple sources, sustainment models can be quickly developed using the best available information.

A case study involving a commercial off-the-shelf fixed-wing aircraft—used and maintained by the U.S. military for decades—demonstrates the architecture’s adaptability across the system lifecycle. The new M&S framework supports dynamic model design and integration with evolving data inputs, enabling scalable, timely, and insightful trade-off analyses that meet the tight schedules encountered by program managers.

Physics simulation model have become essential tools in Amazon’s fulfillment network for analyzing and optimizing material handling equipment (MHE) design. This paper presents a modular framework developed using Emulate3D to model MHE across multiple Amazon fulfillment facilities. The framework incorporates three main components: material to be handled module, material flow control module and package chute module. Through comprehensive validation against pilot results and statistical analysis, the model demonstrates high fidelity in replicating real-world operational dynamics. The framework’s applications span from future automated conveyance flow control logic optimization to existing chute design retrofits. The physics model is used to identify chute jams before launch, location of sensors, optimal jam free merge logic, reducing time/cost for development. Key challenges including package softness, human operator behavior representation, and unexpected operational events are discussed. The study concludes by highlighting future directions in physics simulation to enhance simulation fidelity.

The steel industry’s transformation toward sustainable production leads to an increased use of steel scrap to produce recycled steel. This necessitates a redesign of rail-based supply chains that have been historically designed to serve steel production from coal and iron ore. For a large German rail freight company, our study investigates which shunting yards should potentially be designated as hubs that bundle steel scrap transports. We tackle the complexity of the resulting logistics system with an agent-based simulation model to evaluate the performance of various hub configurations and demand scenarios from a strategic and tactical perspective. Our results show that a smaller number of hubs significantly reduces transport times and operating costs. However, this efficiency comes with increased vulnerability to disruptions, highlighting a trade-off between cost-efficiency and robustness. Our case study offers actionable insights into the efficient and sustainable design of commercial steel scrap transport networks.

The concept of the digital twin has evolved to a key enabler of digital transformation in manufacturing. The adoption of digital twins for factories or digital factory twins remain fragmented and often unclear, particularly for small and medium-sized enterprises. This study addresses this ambiguity by systematically deriving archetypes of digital factory twins to support clearer classification, planning, and implementation. Based on a structured literature review and expert interviews, 71 relevant DFT use cases were identified. The result of the conducted cluster analysis is four distinct archetypes: (1) Basic Planning Factory Twin, (2) Advanced Simulation Factory Twin, (3) Integrated Operations Factory Twin, and (4) Holistic Digital Factory Twin. Each archetype is characterized by specific technical features, data integration levels, lifecycle phases, and stakeholder involvement.

Digital Twins (DT) provide detailed, dynamic representations of production systems, but integrating multiple DTs into a distributed ecosystem presents fundamental challenges beyond mere model interoperability. DTs encapsulate dynamic behaviors, optimization goals, and time management constraints, making their coordination a complex, unsolved problem. Moreover, DT development faces broader challenges, including but not limited to data consistency, real-time synchronization, and cross-domain integration, that persist at both individual and distributed scales. This paper systematically reviews these challenges, examines how current research addresses them, and explores their implications in distributed, hierarchical DT environments. Finally, it presents preliminary ideas for a structured approach to orchestrating multiple DTs, laying the groundwork for future research on holistic DT management.

Modeling and Simulation (M&S) has long been essential for decision-making in complex systems due to its ability to explore strategic and operational alternatives in a structured and risk-free manner. The emergence of Digital Twins (DTs) has further enhanced this by enabling real-time bidirectional synchronization with physical systems. However, constructing and maintaining accurate and adaptive models and DTs remains time- and resource-intensive and requires deep domain expertise. In this paper, we introduce an adaptive Decision-Making Framework (DMF) that integrates Large Language Models (LLMs) and Model-Driven Engineering (MDE) into the M&S and DT pipeline. By leveraging LLMs as proxy experts and synthesis aids, and combining them with MDE to improve reliability, our framework reduces manual effort in model construction, validation and decision space exploration. We present our approach and discuss how it improves agility, reduces expert dependency and can act as a pragmatic aid for making enterprise robust, resilient and adaptive.

Aggregated movement data is widely used for traffic simulations, but privacy constraints often limit data granularity, requiring the use of centroids as cluster representatives. However, centroids might locate cluster centers in contextually irrelevant areas, such as an open field, leading to inaccuracies. This paper introduces a method that leverages an aggregation of points of interest (POIs) such as bus stops or buildings from OpenStreetMap as cluster centers. Using trip data from a suburban region in Germany, we evaluate the spatial deviation between centroids, POIs, and real trip origins and destinations. Our findings show that POI-based centers reduce spatial deviation by up to 46% compared to centroids, with the greatest improvements in rural areas. Furthermore, in an agent-based mobility simulation, POI-based centers significantly reduced travel distances. These results demonstrate that POI-based centers offer a context-aware alternative to centroids, with significant implications for mobility modeling, urban planning, and traffic management.

Academic collaboration is vital for enhancing research impact and interdisciplinary exploration, yet finding suitable collaborators remains challenging. Conventional single-layer random walk methods often struggle with the heterogeneity of academic networks and limited recommendation novelty. To overcome these limitations, we propose a novel Multilayer Random Walk simulation framework (MLRW) that simulates scholarly interactions across cooperation, institutional affiliation, and conference attendance, enabling inter-layer transitions to capture multifaceted scholarly relationships. Tested on the large-scale SciSciNet dataset, our MLRW simulation framework significantly outperforms conventional random walk methods in accuracy and novelty, successfully identifying potential collaborators beyond immediate co-authorship. Our analysis further confirms the significance of institutional affiliation as a collaborative predictor, validating its inclusion. This research contributes a more comprehensive simulation approach to scholar recommendations, enhancing the discovery of latent practical collaborations. Future research will focus on integrating additional interaction dimensions and optimizing weighting strategies to further improve diversity and relevance.

Driven by applications in telecommunication networks, we explore the simulation task of estimating rare event probabilities for tandem queues in their steady state. Existing literature has recognized that importance sampling methods can be inefficient, due to the exploding variance of the path-dependent likelihood functions. To mitigate this, we introduce a new importance sampling approach that utilizes a marginal likelihood ratio on the stationary distribution, effectively avoiding the issue of excessive variance. In addition, we design a machine learning algorithm to estimate this marginal likelihood ratio using importance sampling data. Numerical experiments indicate that our algorithm outperforms the classic importance sampling methods.

High-fidelity simulation models of variable frequency drives often incur expensive computation due to high granularity, complex physics and highly stiff components, hindering real-time Digital Twin Industry 4.0 applications. Surrogate models can outperform simulation solvers by orders of magnitude, potentially making real-time virtual drives feasible within practical computational limits. Despite this potential, current surrogate models suffer from limited generalizability and robustness. In this paper, we present an industrial case study exploring the combination of deep learning with surrogate modeling for simulating variable frequency drives, specifically replacing the induction motor high-fidelity component. We investigate the performance of Long-Short Term Memory-based surrogates, examining how their prediction accuracy and training time vary with synthetic datasets of different sizes, and how well the induction motor surrogates generalize across different motor resistances. This initial study aims to establish a foundation for further development, benchmarking and automation of surrogate modeling workflow for simulation enhancement.

Generation of synthetic data for energy demand allows simulation-based forecasting for infrastructure planning, building optimization, and energy management, key elements of smart cities. This study compares multivariate kernel density estimation (KDE) and time-series generative adversarial networks (TimeGAN) for their ability to generate realistic time series that preserve crucial feature relationships for forecasting. The evaluation is based on both statistical similarity and predictive performance using machine learning models, focusing on seasonal and hourly consumption patterns. The results emphasize the importance of temporal consistency and justify synthetic augmentation when real data is limited, especially for time-aware energy forecasting tasks, and demonstrate how synthesized data can be used when forecasting future energy demand.

This research presents a novel approach to spare parts inventory management by integrating real-time machine health data with dynamic, state-dependent inventory policies. Traditional static models overlook the evolving conditions of industrial machinery. Leveraging advanced digital technologies, such as those pioneered by Augury, our framework dynamically adjusts inventory levels, reducing costs and improving service. Using Markov chain modeling, simulation, and industry collaboration, we demonstrate up to 29% cost savings with state-dependent policies over static base-stock models. Sensitivity analysis confirms the robustness of these strategies.

High responsiveness and economic efficiency are critical objectives in supply chain transportation, both of which are influenced by strategic decisions on shipping mode. An integrated framework combining an efficient simulator with an intelligent decision-making algorithm can provide an observable, low-risk environment for transportation strategy design. An ideal simulation-decision framework must (1) generalize effectively across various settings, (2) reflect fine-grained transportation dynamics, (3) integrate historical experience with predictive insights, and (4) maintain tight integration between simulation feedback and policy refinement. We propose Sim-to-Dec framework to satisfy these requirements. Specifically, Sim-to-Dec consists of a generative simulation module, which leverages autoregressive modeling to simulate continuous state changes, reducing dependence on handcrafted domain-specific rules and enhancing robustness against data fluctuations; and a history–future dual-aware decision model, refined iteratively through end-to-end optimization with simulator interactions. Extensive experiments conducted on three real-world datasets demonstrate that Sim-to-Dec significantly improves timely delivery rates and profit.

This paper introduces a novel method for simulating complex system behaviors using a specific geometric space and force fields within that space. The approach considers the system's performance as a physical trajectory defined by its performance indicators and environmental attributes, which can be deviated by force fields representing risks or opportunities within the system. The primary contribution of this work is the proposal of a method that uses multiple trajectories of a defined system to identify force fields that accurately represent the system

Business Process Simulation (BPS) is crucial for enhancing organizational efficiency and decision-making, enabling organizations to test process changes in a virtual environment without real-world consequences. Despite advancements in automatic simulation model discovery using process mining, BPS is still underused due to challenges in accuracy. Human-in-the-Loop (HITL) integrates human expertise into automated systems, where humans guide, validate, or intervene in the automation process to ensure accuracy and context. This paper introduces a framework identifying key stages in BPS studies where HITL can be applied and the factors influencing the degree of human involvement. The framework is based on a literature review and expert interviews, providing valuable insights and implications for researchers and practitioners.

Identifying suitable collaborators has become an important challenge in research management, where insights into the structure and evolution of scholar collaboration networks are essential. However, existing studies often adopt static or locally dynamic views, limiting their ability to capture long-term network evolution. To address these issues, this paper introduces PENM (Parametric Evolutionary Network Model), a simulation framework designed to model the evolution of scholar collaboration networks through parametric mechanisms. The PENM simulates node and edge evolution through probabilistic rules and tunable parameters, reflecting realistic academic behaviors like cumulative collaboration and co-author expansion. We provide a theoretical analysis of the model's growth patterns under varying parameters and verify these findings through simulation. Evaluations on real-world datasets demonstrate that PENM evolves networks with degree distributions closely aligned with actual scholar networks. PENM offers a versatile simulation-based approach for modeling academic collaboration dynamics, enabling applications and simulation of future academic ecosystems.

Extracting Agent-Based Models (ABMs) from data, also known as Data-Driven Agent-Based Modeling (DDABM), requires a clear understanding of data requirements and their mappings to the corresponding ABM components. DDABM is a relatively new and emerging topic, and as such, there are only highly customized and problem-specific solutions and approaches. In our previous work, we presented a framework for DDABM, identifying the different components of ABMs that can be extracted from data. Building on this, the present study provides a comprehensive analysis of existing DDABM approaches, examining prevailing trends and methodologies, focusing on the mappings between data and ABM components. By synthesizing and comparing different DDABM approaches, we establish explicit mappings that clarify data requirements and their role in enabling DDABM. Our findings enhance the understanding of DDABM and highlight the role of data in automating model extraction, highlighting its potential for advancing data-driven agent-based simulations.

Surgeons’ actions are key to surgical success. Our objective is to develop a decision-support tool to help prioritize patient safety and reduce risks during surgery. We propose a structured mathematical framework that defines key components of a surgical procedure, making it adaptable to various types of surgeries. Using the CholecT50 dataset, we generate and pre-process event logs to construct a process map that models the surgical workflow through Process Mining techniques. This process map provides insights into procedural patterns and can be visualized at different levels of granularity to align with surgeons’ needs. To validate its effectiveness, we simulate synthetic surgeries and assess the process map’s performance in replicating real surgical workflows. By demonstrating the generalizability of our approach, this work paves the way for the development of an advanced decision-support tool that can assist surgeons in real-time decision-making and post-operative analysis.

Simulations are valuable in public health research, with synthetic populations enabling realistic policy analysis. However, methods for generating synthetic populations with domain-specific characteristics remain underexplored. To address this, we previously introduced a population synthesis approach that directly integrates health surveys. This study evaluates its transferability across health outcomes, locations, and timeframes through three case studies. The first generates a Virginia population (2021) with COVID- 19 vaccine intention, comparing results to probabilistic and regression-based approaches. The second synthesizes populations with depression (2021) for Virginia, Tennessee, and New Jersey. The third constructs Virginia populations with smoking behaviors for 2021 and 2022. Results demonstrate the method’s transferability for various health applications, with validation confirming its ability to capture accuracy, statistical relationships, and spatial heterogeneity. These findings enhance population synthesis for public health simulations and offer new datasets with small-area estimates for health outcomes, ultimately supporting public health decision-making.

Coupled Socio-Environmental Networked experiments have been used to represent and analyze complex social phenomena and environmental issues. There is a lack of theory on how to accurately model diverse entities and the connections between them across different spatial and temporal scales. This gap often leads to significant challenges in the modeling, simulating, and analysis of formal experiments. We propose a framework that facilitates software implementation of multi-dimensional coupled socio-environmental networked experiments. Our approach includes: (i) a formal data model paired with a computational model, together providing abstract representations, and (ii) a modeling cycle that maps socio-environmental interactions over time, allowing for multi-action, interactive experiments. The framework is flexible, allowing for a wide variety of network models, interactions, and action sets. We demonstrate its applicability through a case study on agroecological transitions, showing how the modeling cycle and data model can be used to explore socio-environmental phenomena.

Hunger relief networks consist of agencies that work as independent partners within a food bank network. For these networks to effectively and efficiently reduce food insecurity, strategic alliances between agencies are crucial. Agency preference for forming alliances with other agencies can impact network structure and network satisfaction. In this paper, we explore the compatibility and satisfaction achieved by alliances between different agencies. We introduce two agency norms: conservative and diversifying. We develop an agent-based simulation model to investigate alliance formation in a network. We evaluate network satisfaction, satisfaction among different types of agencies, and alliance heterogeneity. We test the statistical significance of satisfaction within a norm and between norms for different agencies. Findings reveal that the ‘diversifying’ norm in the network reduces gaps in satisfaction between strong and weak agencies, ensuring fairness for weaker agencies in the network, whereas the ‘conservative’ norm favors moderate agencies in the network.

Climate change affects thermal comfort and wellness by restricting walkability potential of the built environment. This is especially in the outdoors under the harsh solar radiation exposure of the tropical hot and humid climate. Passive shading strategy plays the most significant role in the walkability potential. Vegetation and man-made structures such as pavements provide shade for comfortable navigation, with the latter being a more sustainable and wellbeing friendly solution. The walkability potential can be simulated using agent-based modelling (ABM) technique. As a heat mitigation strategy to improve the walkability, the most direct intervention is to improve the connectivity of the shading zone along the shortest path between strategic locations. People tend to walk faster and choose the shortest path when dealing with direct sun exposure while avoiding it totally if it gets unbearably hot. The ABM simulation is useful for efficient urban planning of walkability potential in campus.

As the adoption of electric vehicles and hydrogen fuel-cell vehicles grows, understanding how dynamic pricing strategies influence charging and refueling behaviors becomes crucial for optimizing local energy markets. This paper proposes a simulation-based analysis of a hydrogen-electricity integrated charging station that serves both types of vehicles. A multi-agent simulation framework is developed to model the interactions between vehicles and the station, incorporating price- and delay-sensitive behaviors in decision-making. The station can dynamically adjust energy prices, while vehicles optimize their charging or refueling choices based on their utility values. A series of sensitivity analyses are conducted to evaluate how electricity pricing, infrastructure capacity, and waiting behavior impact station performance. Results highlight that moderate electricity prices maximize user participation without sacrificing profit, infrastructure should be right-sized to demand to avoid over- or underutilization, and delay-toleration also affects service outcomes, which may reach the maximum service coverage at the threshold of 45 minutes.

Business enterprises have grappled in the last one and half decade with unavoidable risks of (major) cyber incidents on critical infrastructure (e.g., power grid, cloud systems). The market to manage such risks using cyber insurance (CI) has been growing steadily (but not fast enough) as it is still skeptical of the extent of economic and societal impact of systemic cyber risk across networked supply chains in interdependent IT-driven enterprises. To demystify this skepticism, we study in this paper the role of (a) the statistical nature of multiple enterprise cyber risks contributing to aggregate supply chain risk and (b) the graph structure of the underlying enterprise supply chain network, in the statistical estimation and spread of aggregate cyber risk. More specifically, we provide statistical tail bounds on the aggregate cyber-risk that a cyber risk management firm such as a cyber insurer is exposed to in a supply chain.

The transition toward DC-based industrial grids demands accurate yet efficient planning methods. This paper presents a simulation-driven approach that integrates existing virtual commissioning (VC) models to estimate power demand in early design stages. A minimal-parameter modeling technique is proposed to extract dynamic load profiles from 3D multibody simulations, which are then used for electrical component sizing and system optimization. The methodology is validated using a demonstrator setup consisting of a lift tower and a six-axis industrial robot, both tested under AC and DC operation. Simulation results show close agreement with real measurements, highlighting the method’s ability to capture realistic load behavior with low modeling effort. The approach offers seamless integration into existing planning workflows and supports dimensioning of safe, efficient, and regulation-compliant DC production systems. This contributes to reducing component oversizing, improving energy efficiency, and accelerating the adoption of DC grid architectures in industrial environments.

As the aviation industry works to reduce carbon emissions, airport energy optimization has also been brought into focus. This study explores strategies to reduce peak electricity demand at a major Swedish airport driven by increased electric vehicle charging (EV). EV charging increases grid load, but integrating solar power and battery storage helps stabilize fluctuations and reduce peaks. We present a framework and simulate the combined impact of these factors, demonstrating that smart scheduling with solar and battery systems effectively balances the load. This approach reduces high-load occurrences from 8.6\% to 2.5\%---where 100\% would mean exceeding the threshold year-round---even with 500 additional charging points.

Malaria remains a major global health threat, especially for children under five, causing hundreds of thousands of deaths annually. Proactive Community Case Management (ProCCM) is an intervention designed to enhance early malaria detection and treatment through routine household visits (sweeps), complementing existing control measures. ProCCM is crucial in areas with limited healthcare access and low treatment-seeking rates, but its effectiveness depends on transmission intensity and the coverage of existing interventions. To quantify the impact of ProCCM, we calibrated an agent-based simulation model for settings with seasonal transmission and existing interventions. We evaluated how different ProCCM scheduling strategies perform under varying treatment-seeking rates in reducing severe malaria cases. Our proposed heuristics—greedy and weighted—consistently outperformed a standardized, uniformly spaced approach, offering practical guidance for designing more effective and adaptive malaria control strategies.

Public health experts studying infectious disease spread often seek granular insights into population health outcomes. Metapopulation models offer an effective framework for analyzing disease transmission through subpopulation mixing. These models strike a balance between traditional, homogeneous mixing compartmental models and granular but computationally intensive agent-based models. In collaboration with the Chicago Department of Public Health (CDPH), we developed MetaRVM, an open-source R package for modeling the spread of infectious diseases in subpopulations, which can be flexibly defined by geography, demographics, or other stratifications. MetaRVM is designed to support real-time public health decision-making and through its co-creation with CDPH, we have ensured that it is responsive to real-world needs. We demonstrate its flexible capabilities by tracking influenza dynamics within different age groups in Chicago, by integrating an efficient Bayesian optimization-based calibration approach.

Vaccination is a critical intervention to mitigate the impact of infectious disease outbreaks. However, vaccination decision is complex and influenced by various factors such as individual beliefs, access to vaccines, trust in healthcare systems, and importantly, social norms within communities, the shared understandings and expectations about vaccination behavior. This paper analyzes the impact of social norms on vaccine uptake and subsequent disease transmission by explicitly incorporating these norms into an extension of the agent-based COVID-19 simulation model, COVASIM. We aim to analyze how social norms affect vaccination rates and disease spread. We demonstrate this by implementing community-specific vaccination norms that influence agents through the perceived vaccination behaviors of their social networks. Our simulated case study explored targeted communication about vaccination uptake through different age groups. Through this intervention, we examined the effectiveness of adjusting perceptions of community vaccine uptake to better align with its true value.

This study optimized resource allocation in a standalone Emergency Department projected to experience a 10-30% patient volume increase. Combining data analysis, interviews, and process mapping, a Discrete Event Simulation model was created in Simio, replicating patient flow. The model revealed the ED could manage a 20% volume surge with minor staffing adjustments while maintaining current resources. At 20% increased volume, key metrics such as door-to-provider and treat-and-release times increased to 18 and 200 minutes, surpassing 2023 results by 38% and 12%, respectively. However, exceeding 20% led to an 87% utilization rate for nighttime nurses, creating a potential bottleneck. Minor staffing adjustments mitigated increased treat-and-release times under moderate volume surges, and the site used simulation optimization results to add an 8-hour shift of provider support in the Sunday nighttime hours. This framework offers valuable insights for other EDs anticipating similar challenges, enabling proactive resource management and process optimization.

The increasing prevalence of pediatric mental and behavioral health (MBH) conditions has driven a rise in emergency department (ED) visits, often worsening crowding and straining resources. Psychiatric Emergency Units (PEUs) have emerged as a potential solution to address these challenges by diverting medically stable MBH patients into a calm, specialized setting. We developed a discrete-event simulation of a pediatric ED setting in South Carolina to evaluate the impact of implementing a PEU. Times across various patient journey segments and resource utilization were assessed under varying patient arrival rates and unit capacities. Results showed a shorter length of stay, faster time to disposition and psychiatric evaluation times, and improved room and bed utilization. These findings suggest that PEUs can help hospitals manage increasing MBH volumes more effectively, mitigate system overload, and enhance the quality of pediatric MBH care.

Patients in intensive care units (ICU) require continuous monitoring and care. When physiological abnormalities occur, an alarm is triggered to alert healthcare providers. The alarm response time is influenced by factors such as patient-population profile, care team configuration, staff workload, and unit layout. Response delays can directly impact patient outcomes, emphasizing the need for adequate emergency management capability. While prior simulation studies have explored ICU operations, they often oversimplify dynamic and concurrent interactions between patients and healthcare providers. This study presents a proof-of-concept hybrid simulation model that integrates discrete event simulation (DES) and agent-based simulation (ABS) to comprehensively represent a pediatric ICU environment. By simulating routine activities, patient-triggered alarms, and real-time interactions, the model investigates how varied resource configurations affect response times and outcomes. Built on scalable logic and realistic workflows, the model serves as a foundation for future clinical data integration and supports the development of ICU decision-support applications.

Within hospital pharmacies, aseptic units preparing high-risk injectable medicines face environmental and economic challenges due to resource-intensive processes and emissions. Variability in patient dosage requirements leads to inefficient drug vial usage, resulting in waste generation, carbon emissions generation from waste, and increased costs. Batching could be used to reduce resource consumption and reduce waste associated with single-dose preparation. This study develops a discrete event simulation, as a tool for strategy evaluation and experimentation, to assess the impact of batching on productivity and sustainability. The model captures key process dynamics, including prescriptions arrivals, production processes, and resource consumed. By experimenting with time-sensitive and size-based batching, the study evaluates their effects on the reduction of medical and nonmedical waste, thereby contributing to cost savings, reduction of carbon emissions, and productivity by enhancing workflow efficiency. This study offers insights for hospital pharmacies to evaluate batching strategies effectiveness for reducing waste and promoting sustainability.

This study describes a multi-fidelity simulation framework integrating a high-fidelity discrete event simulation (DES) model with a machine learning (ML)-based low-fidelity model to optimize operating theatre (OT) scheduling in a major public hospital in Singapore. The high-fidelity DES model is trained and validated with real-world data and the low-fidelity model is trained and validated with synthetic data derived from simulation runs with the DES model. The high-fidelity model captures system complexities and uncertainties while the low-fidelity model facilitates policy optimization via the multi-objective non-dominated sorting genetic algorithm (NSGA-II). The optimization algorithm can identify Pareto-optimal policies under varying open access (OA) periods and strategies. Pareto optimal policies are derived across the dual objectives in maximizing OT utilization (OTU) and minimizing waiting time to surgery (WTS). These policies support post-hoc evaluation within an integrated decision support system (DSS).

Reducing waiting times in specialized healthcare has become a pressing concern in many countries, particularly in high-demand services such as traumatology. This study introduces a simulation-based approach to support strategic decision-making for redesigning the referral interface between Primary Care and specialized care, as well as reorganizing internal pathways in the Traumatology Service of the University Hospital of Navarre (Spain). A discrete-event simulation model, developed using real patient data and designed to capture the system’s transient behavior from its current state, is employed to evaluate the effects of these changes on key performance indicators such as number of consultations per patient, physician workload, and waiting list reduction. The model also evaluates how different referral behaviors among Primary Care physicians influence system performance. Results demonstrate the model’s capacity to provide evidence-based guidance for strategic healthcare decisions and highlight its potential to evolve into a digital twin for continuous improvement and operational planning.

In this paper, we consider an outpatient consultation scheduling system with equal-length slots wherein a set of slots each day are reserved for walk-ins. Specifically, we consider the following questions in deciding slot start times to communicate to scheduled patients: (a) should information regarding patient arrival with respect to the slot start time communicated to them (arrival offset with respect to slot start – i.e., are they typically late or early) be considered in deciding the slot start time for communication, and (b) what impact does rounding the slot start time to the nearest 5th or 10th minute have on relevant outcomes? We answer these questions using a validated discrete-event simulation of an FCFS outpatient appointment system in a hospital accommodating both scheduled and walk-in patients. We also describe the development of the simulation itself, which is designed to optimize policies regarding management of walk-in patients and integration of telemedicine.

Formalized models of prenatal care delivery have changed little since 1930, typically including 12-14 in-person visits. Recently released guidelines recommend tailoring prenatal care visit frequency based on patient risk level, potentially integrating telemedicine to increase flexibility for patients. In this paper, we design a discrete event simulation-based approach to study the impact of these new guidelines on healthcare systems, focusing on three main metrics: appointment slot utilization, appointment delays due to lack of capacity, and skipped appointments due to patient cancellations. We find that the tailored approach reduces appointment slot utilization while supporting the same volume of patients as the conventional approach. We also see a decrease in appointment delays and skipped appointments. Our findings suggest that adopting a tailored prenatal care model can reduce provider burnout, allow clinics to accept more patients, and enhance patients’ care experience.

In recent years, hybrid modelling and simulation have become increasingly popular in healthcare for analyzing and improving systems such as patient flow, resource allocation, scheduling, and policy evaluation. These methods combine at least two simulation approaches discrete-event simulation, system dynamics, and agent-based modelling and may also integrate techniques from operations research and management sciences. Their ability to represent complex, dynamic systems has driven their adoption across various healthcare domains. This paper presents a case study of an emergency department (ED) where a hybrid framework combining forecasting models, hybrid simulation, and mixed-integer linear programming was used to optimize physician shift scheduling and improve patient flow and safety. The model outperformed current practices by reducing patient handoffs by 5.6% and decreasing patient time in the ED by 9.2%, without a budget increase. Finally, we propose incorporating reinforcement learning in future work to enable adaptive, data-driven decision-making and further enhance healthcare delivery performance.

This paper is motivated by a panel organized by the Healthcare and Life Sciences track at the 2025 Winter Simulation Conference (WSC). We summarize the panelists' perspectives and reflect on current trends and future research directions for simulation applications in healthcare and life sciences. We begin with a brief review of key methodologies and application trends from the past decade of WSC proceedings. We then present expert insights from a range of application areas, including (bio)pharmaceutical manufacturing, hospital operations, public health and epidemiology, and modeling human behavior. The panelists provide diverse perspectives from academia and industry, and highlight emerging challenges, opportunities, and future research directions to advance simulation in healthcare and life sciences.

Nationwide patient concentration poses a significant burden on healthcare systems, largely due to patients’ perception that metropolitan regions offer superior care quality. To better understand this phenomenon, we present an agent-based model to examine how objective quality (OQ) and perceived quality (PQ) co-evolve in a free-choice healthcare system, using South Korea as a salient case. Four mechanisms — preferential hospital choice, scale effect, quality recognition, and word-of-mouth — form a feedback loop: concentration raises OQ, utilization updates PQ, and perceptions diffuse through the population. We identify three emergent phenomena — local dominance, global dominance, and asymmetric quality recognition — and interpret how each contributes to patient outmigration. Building on these insights, we further explore strategies such as “local tiering” and “information provision.” This model-based approach deepens understanding of OQ–PQ dynamics, and offers insights for addressing nationwide healthcare utilization in various contexts.

Liver transplantation is the second most common transplant procedure in the United States and the only curative treatment for patients with end-stage liver disease, which is one of the leading causes of death nationwide. The United Network for Organ Sharing operates the national liver transplant waiting list and allocates organs under a complex priority system based on medical urgency, geography, and waiting time. Healthcare providers accept or refuse liver offers based on transplant candidates’ medical needs and donor quality, among other factors. We develop a simulation environment to assess current acceptance practices based on a Markov reward process. Our simulation framework models organ arrivals and patients' health progression as continuous-time processes and mimics how decisions are made in practice using a randomized policy. Based on our simulation framework, we provide insights and identify areas for enhancing patient management and liver offer acceptance.

Biopharmaceutical drugs have transformed modern medicine, yet their manufacturing processes remain challenged by yield variability and rising production costs. This paper explores a promising application in a new domain to improve biomanufacturing efficiency through the \textit{backpass} production method. Our production setting consists of two bioreactors: a dedicated bioreactor for production of a high-value product A, and a shared bioreactor for product A and a lower-value product B. The backpass method allows biomass to be transferred from the dedicated to the shared bioreactor, enabling additional production of product A while bypassing extensive upstream processing steps. We examine the performance of three backpass strategies, defined based on feedback from our industry partner, and use discrete event simulation to answer industry-specific questions related to the system's performance regarding throughput and profitability. This analysis provides practitioners with a decision-support framework for capital investments and operational planning.

Orthopedic health services are characterized by high patient volumes, long elective waits, unpredictable emergency demand, and close coupling with other hospital processes. These present significant challenges for meeting operational targets and maintaining quality of care. In healthcare, simulation has been widely used for addressing similar challenges. This systematic scoping review identifies and analyzes academic papers using simulation to address operational-level challenges for orthopedic service delivery. We analyzed 37 studies over two decades, combining a structured analysis with topic modelling to categorize and map applications. Despite widespread recognition of its potential, simulation remains underutilized in orthopedics, with fragmented application and limited real-world implementation. Recent trends indicate a shift toward system-wide approaches that better align with operational realities and stakeholder needs. Future research should aim to bridge methodological innovation with collaboration and practical application, such as hybrid and real-time simulation approaches focusing on stakeholder needs, and integrating relevant operational performance metrics.

Precision medicine (PM) is an approach that aims to tailor treatments based on patient profiles (patients' biometric characteristics). In PM practice, treatment performance is typically evaluated through simulation models or clinical trials. Although these two methods have differences in their sampling subjects and requirements, both are based on a sequential sampling process and require determining a stopping time for sampling to ensure that, with a prespecified confidence level, the best treatment is correctly identified for each patient profile. In this research, we propose unified stopping rules applicable to both simulation and clinical trial-based PM sampling processes. Specifically, we adapt the generalized likelihood ratio (GLR) test to determine when samples collected are sufficient and calibrate it using mixture martingales with a peeling method. Our stopping rules are theoretically grounded and can be integrated with different types of sampling strategies. Numerical experiments on synthetic problems and a case study demonstrate their effectiveness.

Despite positive clinical outcomes of Chimeric Antigen Receptor T-cell therapy, its time-sensitivity causes substantial logistical challenges. This paper introduces a simulation-optimization (SO) framework to address both transportation mode selection and patient scheduling in CAR T-cell therapy supply chains. This framework combines a discrete-event simulation with a metaheuristic optimization to handle uncertainties in processing times and patient conditions. Experiments across different time-window constraints demonstrate that the developed SO model consistently outperforms traditional scheduling heuristics (FIFO, SPT, EDD) in total cost while maintaining timely delivery. This model provides a superior balance between transportation efficiency and delay minimization compared to rule-based methods. Results highlight the potential of simulation-based optimization to enhance personalized medicine delivery by improving cost-effectiveness without compromising treatment timeliness.

Non-pharmaceutical interventions (NPIs) are the immediate public health reaction to emerging epidemics. While they generally help slow down infection dynamics, they can be associated with relevant socioeconomic costs, like lost school- or work days caused by preemptive household quarantines. However, research suggests that not all households contribute equally to the overall infection dynamics. In this study, we introduce the novel “Infection Contribution” metric that allows us to trace the involvement of particular household types over entire infection chains. Building upon the German Epidemic Microsimulation System, we quantify the impact of various household types, considering their size and composition in a COVID-19-like scenario. Additionally, we show how targeting interventions based on household characteristics produces efficient strategies, outperforming non-selective strategies in almost all scenarios. Our approach can be transferred to other NPIs, such as school closure, testing, or contact tracing, and even inform the prioritization of vaccinations.

The growing complexity of public health emergencies requires modeling tools that are both scientifically robust and operationally scalable. As part of the EU Horizon 2020 STAMINA project, we deployed the Flu and Coronavirus Simulator (FACS), a geospatial agent-based model designed to simulate the spread of infectious diseases at local and regional levels. This paper presents a case study from a UK Public Health Workshop, where FACS supported the evaluation of epidemic response scenarios. We describe how FACS integrates demographic, spatial, and epidemiological data, and outline key enhancements, such as location-based parallelization and FabSim3-enabled automation, which enable large-scale simulation. We detail the scenario designs and outcomes, highlighting the intersection of simulation projections and intervention planning. Finally, we reflect on communicating results to stakeholders and bridging the gap between modeling and policy. This work demonstrates how geospatially grounded, scalable agent-based simulations can provide meaningful insights into regional intervention planning within operational timeframes.

As AI integration in critical domains grows, evaluating its effectiveness in complex policy environments remains challenging. We introduce a two-stage simulation framework for assessing AI policy recommendations in the COVID-19 pandemic. First, we train a deep reinforcement learning (DRL) agent using data from 186 countries to model optimal intervention timing and intensity. Results suggest the DRL agent outperforms average government outcomes within our simplified model under specific assumptions, reducing infections and fatalities, improving recovery rates. Second, we employ SEIRD (Susceptible-Exposed-Infected-Recovered-Dead) modeling to create a dynamic simulation environment, testing the agent across diverse scenarios beyond historical data. Unlike prior work lacking systematic evaluation, our framework provides a controlled testbed for high-stakes policy decisions before implementation. This presents a responsible approach to AI evaluation where real-world experimentation raises ethical concerns. It highlights the role of simulations in bridging development-deployment gaps while identifying financial constraints and human-AI interaction as future research priorities.

The Digital Twin (DT) concept has garnered significant interest in recent years, with its potential to improve system efficiency through data synchronization and model updates. Unlike traditional modeling techniques, DTs dynamically incorporate operational data to simulate "what-if" scenarios. However, the growing number of definitions and applications of DTs across industries highlights the need for a unified theoretical framework. This paper proposes a system-theoretic approach to formalize the concept of DT, aiming to contribute to a comprehensive DT theory. The paper introduces the DMSµ framework and explores the potential of hierarchical hybrid automata to offer within that framework a formalism that supports symbolic manipulation. It provides a foundation for advancing the formalization and application of DT technology across diverse fields.

One of the primary in-built components of smart, continuous manufacturing lines is the production speed management system (PSMS). In addition to being overly cautious, the decisions made in these systems may center on making local adjustments to the manufacturing process, indicating a major drawback of such systems that prevents them from acting as proper digital twins. This study delves into hybridizing the continuous and discrete event simulation, DOE, and V-graph methods to redefine PSMS’s internal decision algorithms and procedures, giving it an aerial perspective of the line and turning it into a stand-alone online digital twin with decisions at a system level. The proposed approach is applied to a practical case from the food and beverage industry to validate its effectiveness. Numerical results demonstrated an intelligent, dynamic balancing of the production line, a substantial increment in productivity, and up to 37.7% better resiliency against new failure and repair patterns.

Managing production, inventory, and demand forecasting in a competitive market is challenging due to consumer behavior and market dynamics. Inefficient forecasting can lead to inadequate inventory, an interrupted production schedule, and eventually, less profit. This study presents a simulation-based decision support framework integrating discrete event simulation (DES) and system dynamics (SD). DES models production and inventory management to ensure optimized resource utilization, while SD is employed to incorporate market dynamics. This model jointly determines demand through purchase decisions from potential users and replacement demand from existing adopters. Further refinements prevent sales declines and sustain long-term market stability. This hybrid simulation approach provides insights into demand evolution and inventory optimization, aiding strategic decision-making. Finally, we propose and integrate a dynamic marketing strategy algorithm with the simulation model, which results in around 38% more demand growth than the existing demand curve. The proposed approach was validated through rigorous experimentation and optimization analysis.

Traffic-vehicle co-simulation couples microscopic traffic simulation with full-body vehicle dynamics to assess system-level impacts on mobility, energy, and safety with greater realism. Incorporating elevation is critical for accurately modeling vehicle behavior and energy use, especially for gradient-sensitive vehicles such as electric and heavy-duty trucks. However, raw elevation data often contain noise, discontinuities, and inconsistencies. While such issues may be negligible in traditional traffic simulations, they significantly affect traffic-vehicle co-simulations where vehicle dynamics are sensitive to road grade variations. This paper investigates the impact of unprocessed elevation data on vehicle behavior and energy consumption using a 42-mile simulation along Interstate 81. To mitigate these effects arising from elevation data issues, we propose an elevation processing workflow to improve the realism and stability of traffic-vehicle co-simulation. Results show that the method effectively removes noise and abrupt elevation transitions while preserving roadway geometry.

Ensuring traffic safety remains a major challenge due to the complexity of traffic environments and the early stage of autonomous vehicle (AV) technology, despite their potential to significantly reduce accidents and enhance road safety. The Artificial Potential Field (APF) approach offers a promising solution by simulating how vehicles adjust their motion, speed, and interactions with surrounding vehicles to maintain safety. This paper aims to introduce a conceptual hybrid simulation using the APF implemented within a multi-agent framework. The objective is to evaluate the suitability of APF model for real-time safety applications across extended time periods and diverse traffic scenarios. This evaluation is conducted through a hybrid simulation approach to identify advantages and limitations compared to existing risk assessment methodologies.

The rise of ride-sourcing services has changed the transportation industry, reshaping urban mobility services. This paper presents an integrated framework of the adoption of ride-sourcing services and its impact on transportation markets using a combined approach of System Dynamics (SD) and Extended-Decision Field Theory (E-DFT). Drawing on data from ride-sourcing platforms such as Uber and Lyft, the study investigates the temporal dynamics and trends of ride-sourcing demand. SD modeling is employed to capture the complex interactions and feedback loops within the ride-sourcing ecosystem at system-level. The integration of System Dynamics and extended DFT allows for a more comprehensive and holistic modeling of the ride-sourcing market. It enables exploration of various scenarios and policy interventions, providing insights into the long-term behavior of the market and facilitating evidence-based decision-making by policymakers and industry stakeholders while accommodating individual users' decisions based on changing preferences and environments.

The healthcare sector faces rising demands from aging populations, complex patient needs, and workforce shortages, requiring innovative solutions for resource and service coordination. In response, Trondheim municipality and St. Olavs hospital are developing integrated planning tools within the Research Council of Norway funded HARMONI project. This transdisciplinary initiative aims to bridge the widening gap between service demand and capacity through planning tools that offer an overview of patient flows and resource capacities. The tools will support comprehensive planning, process adjustments, and capacity dimensioning, fostering collaboration to prevent patient queues and unnecessary transfers. We present three ongoing hybrid model case studies focusing on patient flow between primary and specialist health services. These cases provide opportunities to develop hybrid simulation verification and validation techniques.

Colorectal cancer (CRC) is the third leading cause of cancer-related death in the U.S., despite being largely preventable through screening. Interventions such as mailing fecal immunochemical tests (FIT) or sending patient reminders have shown varying success in increasing screening rates. Simulation modeling has played a key role in estimating the impact of these interventions on long-term health outcomes by representing the natural history of CRC. Using a simulation model of CRC, in this work, we developed a metamodeling-based decision tree to help clinics and health systems select CRC screening interventions that best match their population. Our approach uses estimates of intervention effectiveness based on pre-intervention screening levels, eliminating the need for users to assume how an intervention will impact outcomes. By tailoring recommendations to population characteristics and baseline screening rates, the decision tree supports data-driven decisions to improve CRC screening and, ultimately, population health.

Antibiotic resistance (ABR) is a major global health threat, contributing to increased mortality and economic losses. The emergence and spread of ABR is driven by healthcare practices, environmental contamination, and human-to-human transmission. In low and middle income countries (LMICs), limited healthcare infrastructure and environmental factors, such as contaminated water and poor sanitation, exacerbate the situation. In these regions, inappropriate antibiotic use and insufficient infection control measures further promote resistance. This paper presents a hybrid model that combines System Dynamics (SD) and Agent-Based Modeling (ABM) to explore complex interactions between healthcare systems, environmental factors, and human behavior in community settings. The SD approach models aggregated within-host bacterial dynamics and external environmental factors, while ABM captures individual person behaviors community interactions and interactions with healthcare. By integrating these methods, this study offers a more comprehensive framework for understanding the emergence of ABR in LMICs. Preliminary results and future directions are discussed.

Although Large Language Models get a lot of attention, Generative Artificial Intelligence encompasses a variety of methods such as Generative Adversarial Networks, Variational Autoencoders or Diffusion Models, that all work very differently but are all capable of generating synthetic data. These methods have considerable potential to make simulation studies more efficient, especially through the creation of artificial data sets, automatic model parameterization and assisted result analysis. The aim of this study is to systematically classify generative methods and their applicability in the context of simulation studies. Based on a comprehensive literature review, applications, trends and challenges of generative methods that are used in combination with simulation are analyzed and structured. This is then summarized in a conceptual workflow that shows how and in which phase generative methods can be used advantageously in simulation studies.

We explore the role of ontologies in enhancing hybrid modeling and simulation through improved semantic rigor, model reusability, and interoperability across systems, disciplines, and tools. By distinguishing between methodological and referential ontologies, we demonstrate how these complementary approaches address interoperability challenges along three axes: Human–Human, Human–Machine, and Machine–Machine. Techniques such as competency questions, ontology design patterns, and layered strategies are highlighted for promoting shared understanding and formal precision. Integrating ontologies with Semantic Web Technologies, we showcase their dual role as descriptive domain representations and prescriptive guides for simulation construction. Four application cases – sea-level rise analysis, Industry 4.0 modeling, artificial societies for policy support, and cyber threat evaluation – illustrate the practical benefits of ontology-driven hybrid simulation workflows. We conclude by discussing challenges and opportunities in ontology-based hybrid M&S, including tool integration, semantic alignment, and support for explainable AI.

Large Language Models (LLMs) offer transformative potential for Modeling & Simulation (M&S) through natural language interfaces that simplify workflows. However, over-reliance risks compromising quality due to ambiguities, logical shortcuts, and hallucinations. This paper advocates integrating LLMs as middleware or translators between specialized tools to mitigate complexity in M&S tasks. Acting as translators, LLMs can enhance interoperability across multi-formalism, multi-semantics, and multi-paradigm systems. We address two key challenges: identifying appropriate languages and tools for modeling and simulation tasks, and developing efficient software architectures that integrate LLMs without performance bottlenecks. To this end, the paper explores LLM-mediated workflows, emphasizes structured tool integration, and recommends Low-Rank Adaptation-based architectures for efficient task-specific adaptations. This approach ensures LLMs complement rather than replace specialized tools, fostering high-quality, reliable M&S processes.

Existing manufacturing research on greenhouse gas emissions often focuses on Scope 1 and Scope 2 emissions and underestimates Scope 3 emissions, which are indirect emissions from a firm’s value chains, city and region consumption. Traditional methodologies for evaluating carbon emissions are limited for Scope 3 emissions, due to the complexity of manufacturing supply chains and lack of quality data, leading to incomplete carbon accounting and potential double-counting. This challenge is pronounced for high value manufacturing, an emergent manufacturing perspective, due to the complexity of its supply chain network. This study develops a comprehensive hybrid modeling framework for evaluating Scope 3 emissions at product level, useful for manufacturers and modelers.

We introduce a hybrid modeling approach for simulating carbon dioxide (CO2) dispersion in indoor environments by integrating Cellular Automata (CA), Discrete Event System Specification (DEVS), and agent-based modeling. The proposed framework enhances traditional models by incorporating dynamic CO2 generators, random-walk algorithms, and CO2 sinks. We show how the method can be used to examine the effects that room layouts, occupant movement, ventilation settings, and CO2 sinks and sources placement have on indoor concentration patterns. The approach presented here enables the exploration of various configuration parameters and provides a flexible and scalable tool for understanding CO2 diffusion.

The paper deals with the application of hybrid simulation in the field of socio-economic systems. We present a model for assessing the extent to which interruptions in work during a woman's professional career affect the amount of her future pension. Breaks in work due to the childbirths, their upbringing, as well as shorter breaks due to emergency care of sick children significantly affect the amount of the first pension collected in defined contribution pension systems. The model is a hybrid of a demographic model developed according to the system dynamics approach and a pension model built with discrete event simulation paradigm. The experiments examine the size of the maternity penalty for different career scenarios and long-term demographic changes. The primary research objective was to examine the extent to which women's individual life-course choices affect their pension outcomes, in interaction with systemic features of pension schemes and projected demographic changes.

The Kotlin Simulation Library (KSL) is an open-source library written in the Kotlin programming language that facilitates Monte Carlo and discrete-event simulation modeling. The library provides an API framework for developing, executing, and analyzing models using both the event view and the process view modeling perspectives. This paper provides a tutorial on modeling with resources within simulation models. The KSL will be utilized to illustrate important concepts that every simulation modeler should understand within the context of modeling resources within a simulation model. A general discussion of resource modeling concepts is presented. Then, examples are used to illustrate how to put the concepts into practice. While the concepts will be presented within the context of the KSL, the ideas should be important to users of other simulation languages. This tutorial provides both an overview of resource modeling constructs within the KSL and presents tutorial examples.

Conceptual modeling entails the abstraction of a simulation model from the system of interest to create a simplified representation of the real world. This activity is vital to successful simulation modeling, but it is not well understood. In this tutorial we aim to develop a better understanding of conceptual modeling by exploring this activity from various perspectives. Our initial focus is on defining both a conceptual model and the activity of conceptual modeling. The activity is further explored from a practice-based perspective with reference to an example that is based on actual events. Conceptual modeling is then discussed from three further perspectives: understanding the relationship between a simulation model’s accuracy and its complexity; the role of assumptions and simplifications in modeling; and frameworks for guiding the activity of conceptual modeling. This exploration of the concepts of conceptual modeling provides the underlying knowledge required for improving our conceptual models for simulation.

This tutorial paper explores the applicability of Generative AI (GenAI), particularly Large Language Models (LLMs), within the context of simulation modeling. The discussion is organized around three key perspectives: Generation, where GenAI is used to create simulation models and input data; Execution, where it supports real-time decision-making and adaptive logic during simulation runs; and Analysis, where GenAI assists in running experiments, interpreting results, and generating insightful reports. In addition to these core perspectives, the paper also covers practical implementation considerations such as prompt engineering, fine-tuning, and Retrieval-Augmented Generation (RAG). The tutorial offers both conceptual guidance and hands-on examples to support researchers and practitioners seeking to integrate GenAI into simulation environments.

The adoption of data-driven Digital Twins in smart manufacturing systems necessitates robust, data-driven modeling techniques. Stochastic Petri Nets (SPNs) offer a formal framework for capturing concurrency and synchronization in discrete event systems, making them well-suited for modeling of smart manufacturing systems. This tutorial provides a hands-on introduction to extracting SPN models from system event logs. Using our Python-based SPN library (PySPN), we guide participants through generating ground-truth models in Petri Net Markup Language (PNML), simulating SPNs for generating event logs, and applying Process Mining techniques for automated SPN extraction. Two case studies illustrate the end-to-end workflow, including SPN model validation for Digital Twins. By the end of the tutorial, participants will gain practical skills in data-driven SPN modeling for Digital Twins applications in manufacturing.

In this tutorial we give a definitive and comprehensive 10-step approach for conducting a successful simulation study. Topics to be discussed include problem formulation, collection and analysis of data, developing a valid and credible model, modeling sources of system randomness, design and analysis of simulation experiments, model documentation, and project management.

Digital twins are rapidly emerging as a disruptive innovation in industry, offering a dynamic integration of physical systems with their virtual counterparts through data-driven and real-time synchronization. In this work, we explore the foundational principles, architectures, and life cycle of digital twins, with a particular emphasis on their use in simulation-based decision-making. We dissect the core components of a digital twin and examine how these elements interact across a system’s life cycle. The notion of a twinning system is introduced as a unifying framework for sensing, simulation and digital models/shadows/twins, with applications in diverse fields. We outline the most significant choices that stakeholders must make. A case-study is based on a lab-scale manufacturing system and illustrates how DTs are constructed, validated, and used for scenario analysis and autonomous decision-making. We also discuss the challenges associated with synchronization, model validity, and the development of internal services for operational use.

Deep Reinforcement Learning (DRL) has shown considerable promise in addressing complex sequential decision-making tasks across various fields, yet its integration within Operations Research (OR) remains limited despite clear methodological compatibility. This paper serves as a practical tutorial aimed at bridging this gap, specifically guiding simulation practitioners and researchers through the process of developing DRL environments using Python and the Gymnasium library. We outline the alignment between traditional simulation model components, such as state and action spaces, objective functions, and constraints, and their DRL counterparts. Using an inventory control scenario as an illustrative example, which is also available online through our GitHub repository, we detail the steps involved in designing, implementing, and integrating custom DRL environments with contemporary DRL algorithms.

This work presents substantial improvements to Enhance, a recent approach for graph partitioning in large-scale distributed microscopic traffic simulations, particularly in challenging load-balancing scenarios within heterogeneous computing environments. With a thorough analysis of the diffusive and refinement phases of the Enhance algorithm, we identified orthogonal opportunities for optimizations that markedly improved the quality of the generated partitionings. We validated these improvements using synthetic scenarios, achieving up to a 46.5\% reduction in estimated runtime compared to the original algorithm and providing sound reasoning and intuitions to explain the nature and magnitude of the improvements. Finally, we show experimentally that the performance gains observed in the synthetic scenario partially translate into performance gains in the real system.

This research aimed to develop a methodology and a framework to calibrate microscopic simulation models driving behaviors to reproduce safety conflicts observed in real-world environments. The Intelligent Driver Model (IDM) was selected as the car-following algorithm to be utilized in the External Driver Model (EDM) Application Programming Interface (API) in VISSIM to better represent real-world driving behavior. The calibration method starts with an experiment design in the statistical software JMP Pro 16, that provided 84 simulation runs, each with a distinct combination of the 11 EDM input variables. After 84 runs with such variables, the traffic trajectory was analyzed by the FHWA’s Surrogate Safety Assessment Model (SSAM) to generate crossing, rear-end, and lane change conflict counts. It is concluded that the calibration method proposed can closely match the conflict counts translated from real-world conditions.

Electric delivery vehicles are becoming increasingly common in last-mile logistics. Thus, the shift to electric transport introduces new challenges to urban transportation management, as these vehicles are range-constrained under heavy loads and require charging infrastructures on a daily basis. Battery systems constitute the most critical component of these vehicles and understanding their energy consumption is essential. This paper presents a simulation-based framework to evaluate battery performance under realistic delivery scenarios, focusing on how operational variables, such as speed, payload, and travel distance affect power consumption. Hence, an integrated agent-based simulation model is proposed, incorporating an equivalent electric circuit model of the battery, and a mechanical model of the vehicle energy requirements. Moreover, ground-based vehicles are considered for experiments. Two battery configurations are used in the simulated scenarios, representing small and medium-sized battery systems. Similarly, short and long delivery times are considered to evaluate the impact of battery size on distribution.

This study explores the integration of machine learning (ML) clustering techniques into a simulation-optimization framework aimed at enhancing the efficiency of airport security checkpoints. Simulation-optimization is particularly suited for addressing problems characterized by evolving data uncertainties, necessitating critical system decisions before the complete data stream is observed. This scenario is prevalent in airport security, where passenger arrival times are unpredictable, and resource allocation must be planned in advance. Despite its suitability, simulation-optimization is computationally intensive, limiting its practicality for real-time decision-making. This research hypothesizes that incorporating ML clustering techniques into the simulation-optimization framework can significantly reduce computational time. A comprehensive computational study is conducted to evaluate the performance of various ML clustering techniques, identifying the OPTICS method as the best found approach. By incorporating ML clustering methods, specifically the OPTICS technique, the framework significantly reduces computational time while maintaining high-quality solutions for resource allocation.

Aircraft manufacturing presents significant challenges for logistics departments due to the complexity of processes and technology, as well as the high variety of parts that must be handled. To support the development and optimization of these complex logistics processes in the aviation industry, simulation is commonly employed. However, existing simulation models are typically tailored to specific use cases. Reusing or adapting these models for other aircraft-specific applications often requires substantial implementation and validation efforts. As a result, there is a need for flexible and easily adaptable simulation models. This work aims to address this challenge by developing a modular library for logistics processes in aircraft manufacturing. The outcome of this work highlights the simplifications introduced by the developed library and its application in a real aviation warehouse.

This paper introduces a methodology for simulating flight dynamics utilizing the Tool Command Language (Tcl). Tcl, created by John Ousterhout, was conceived as an embeddable scripting language for an experimental Computer Aided Design (CAD) system. Tcl, a mature and maturing language recognized for its simplicity, versatility, and extensibility, is a compelling contender for the integration of flight dynamics functionalities. The work presents an extension method utilizing Tcl's adaptability for a novel type of flight simulation programming. Initial test findings demonstrate performance appropriate for the creation of human-in-the-loop real-time flight simulations. The possibility for efficient and precise modeling of future complicated distributed simulation elements is discussed, and recommendations regarding subsequent development priorities are drawn.

This study presents an agent-based modeling framework for enhancing the efficiency and sustainability of perishable food supply chains. The framework integrates forward logistics redesign, reverse logistics, and waste valorization into a spatially explicit simulation environment. It is applied to the tomato supply chain in Jordan, restructuring the centralized market configuration into a decentralized closed loop system with collection points, regional hubs, and biogas units. The model simulates transportation flows, agent interactions, and waste return through retailer backhauls. Simulation results show a 31.1 percent reduction in annual transportation distance and cost, and a 35.9 percent decrease in transportation cost per ton. The proposed approach supports cost-effective logistics and a more equitable distribution of transport burden, particularly by shifting a greater share to retailers. Its modular structure, combined with reliance on synthetic data and scenario flexibility, makes it suitable for evaluating strategies in fragmented, resource-constrained supply chains.

Open access to electric vehicle charging session data is limited to a small selection provided by operators of mostly public or workplace chargers. This restriction poses a hurdle in research on regional energy demand shift flexibilities enabled by smart charging, since usage characteristics between different charging options remain hidden. In this paper, we present an agent-based simulation model with parameterizable availability and usage preferences of public and private charging infrastructure to access insights of charging behavior otherwise only visible through proprietary data. Thus, we enable utility operators to estimate spatial charging energy distribution and support the integration of renewable energy by showing potentials for smart charging. In a first application, we point out how increased access and use of private charging facilities can lead to additional energy demand in rural municipalities, which, in turn, leads to a lower grid load in urban centers.

Recent advances allow Less-Than-Truckload (LTL) terminals to know the distribution of arriving goods on inbound trucks in advance. Therefore, assigning docks to inbound and outbound relations and trucks to docks is a critical problem for terminal operators. This paper introduces a framework combining automatic model generation and optimization. The approach aims to allow testing of suggestions from multiple optimization algorithms. The relevance and feasibility of this approach in finding an appropriate optimization algorithm for a given system are demonstrated through a simplified case study of a variation of the dock assignment problem. This paper demonstrates how such a combination can be constructed and how the methods can effectively complement each other, using the example of LTL terminals.

Volatile customer demand poses a significant challenge for the logistics networks of trading companies. To mitigate the uncertainty in future customer demand, many products are produced to stock with the goal to be able to meet the customers’ expectations. To adequately manage their product inventory, demand forecasting is a major concern in the companies’ sales planning. A promising approach besides using observational data as an input for the forecasting methods is simulation-based data generation, called data farming. In this paper, purposeful data generation and large-scale experiments are applied to generate input data for predicting customer demand in sales planning of a trading company. An approach is presented for using data farming in combination with established forecasting methods such as random forests. The application is discussed on a real-world use case, highlighting benefits of the chosen approach, and providing useful and value-adding insights to motivate further research.

The growing trend of specialization is significantly increasing the importance of Electronic Manufacturing Services (EMS) providers. Typically, EMS companies operate within global supply networks characterized by high complexity and dynamic interactions between multiple stakeholders. As a consequence, EMS providers frequently experience volatile and opaque procurement and production planning processes. This paper investigates the potential of collaborative planning between EMS providers and their customers to address these challenges. Using discrete-event simulation, we compare traditional isolated planning approaches with collaborative planning strategies. Based on empirical data from an EMS company, our findings highlight the benefits of collaborative planning, particularly in improving inventory management and service levels for EMS providers. We conclude by presenting recommendations for practical implementation of collaborative planning in the EMS industry.

Automated guided vehicles (AGVs) and Autonomous mobile robots (AMRs) are widely used in intralogistics for material delivery and product transport. As manufacturing evolves with Industry 5.0, the integration of autonomous systems, particularly in complex shop floor layouts, plays a crucial role in improving efficiency and reducing human intervention. This study explores various intralogistics concepts for AGVs and AMRs collaborating on an assembly line. We define an online dispatching rule for AGVs and benchmark it against previous implementations on high-mix-low-volume production. All proposed concept-related decisions are analyzed via a simulation model under a real-world case study, with sensitivity analyses varying system parameters. We observe an improvement of up to 51% in the hourly throughput rate and a significant drop in AGV utilization rates of up to 88% compared to benchmark instances. The results also show that pooling AGV/AMR resources achieves a higher throughput rate, and consequently reduces investment costs.

Industry 4.0 forces a significant transition in the field of manufacturing and production planning and control. A corner stone of Industry 4.0 scenarios is the ability of control systems to adapt autonomously to changes on the shop floor. Reinforcement Learning is considered as an approach to achieve this target. Consequently, the control strategies of the agents trained by Reinforcement Learning need to generalize in a manner that the agents are able to control modified production systems they are not directly trained for. This paper presents an evaluation of the generalization properties of a decentralized multi-agent Reinforcement Learning algorithm for controlling flow shops using complex gantry robot systems for automated material handling. It is shown that the corresponding agents are able to cope with variations of the production and gantry robot system, as long as these variations are within realistic boundaries, and thus are suitable for Industry 4.0 scenarios.

Decision Support Systems (DSS) are a crucial component in production logistics, aiding companies in solving complex decision problems with multiple influences. This publication provides a structured review of the literature on the application of DSS in production logistics, focusing on the applied methods for decision support, such as optimization or Artificial Intelligence. The analysis considers scientific publications from 2015 to 2024, including industry use cases. Data analysis of categorizations of DSS is used. The findings highlight trends and limitations of current DSS application cases from the literature. Optimization methods, particularly heuristic and metaheuristic, are the most commonly employed decision support methods, followed by simulation. Despite the increased interest in AI technologies, their role in DSS for production logistics remains secondary. Like simulation methods, AI technologies are highly relevant when combined with optimization methods. The study provides a foundation for future research and practical advancements in decision support for manufacturing environments.

Reinforcement Learning (RL) has evolved as a dominant AI method to move towards optimal control of complex systems. In the domain of manufacturing, production control has emerged as one of the major application areas, where the material flow is governed by an RL agent that is fed with the real-time information flow of the system through a digital twin. A digital twin framework facilitates efficient and effective RL training. The coordination task performed in discrete, flexible manufacturing offers multiple decisions that cannot be found in all cases. Hoist scheduling for galvanic equipment introduces additional constraints as parts cannot be left inside a galvanic bath arbitrarily long. Even short deviations critically affect product quality, which is even more complicated in the mix high volume environments. The proposed RL agent learns superior control compared to the state-of-the-art and simple heuristic rules are derived for everyday application in the absence of digital twins.

Streamlining the order release strategy for a job-shop production system under uncertainty is a complex problem. The system is likely to have a number of stochastic parameters contributing to the problem complexity. These factors make it challenging to develop optimal job-shop schedules. This paper presents a Reinforcement Learning-based discrete-event simulation approach that streamlines the policy for releasing orders in a job-shop production line under uncertainty. A digital twin (DT) was developed to simulate the job-shop production line, which facilitated the collection of process and equipment data. A reinforcement learning algorithm is connected to the DT environment and trains with the previously collected data. Once the training is complete, its solution is evaluated in the DT using experimental runs. The method is compared with a few popular heuristic-based rules. The experimental results show that the proposed method is effective in streamlining the order release in a job-shop production system with uncertainty.

Production Planning and Control (PPC) faces increasing complexity due to volatile demand, high product variety, and dynamic shop floor conditions. Reinforcement Learning (RL) offers adaptive decision-making capabilities to address these challenges. RL often relies on simulation environments for the intensive training, allowing for short run times during execution. This paper reviews existing literature to examine how RL agents are modeled in terms of state space, action space, and reward function, focusing on order release and related production scheduling tasks. The findings reveal considerable variation in modeling approaches and a lack of theoretical guidance, particularly in reward design and feature selection.

Simulation has been an indispensable tool for the design, analysis, and control of manufacturing systems for decades. With new digital twinning technologies and artificial intelligence capabilities appearing in a fast pace, will simulation -- as we know today -- retain its prominent role in the manufacturing industry of the future? What are the current and projected trends in manufacturing industry that will make simulation even more relevant? How will simulation evolve to address the needs of next-generation factories? Motivated by these initial questions, the Manufacturing and Industry 4.0 track of 2025 Winter Simulation Conference (WSC) brings together a panel of experts to discuss the future of simulation in manufacturing.

The concept of the digital twin has evolved to a key enabler of digital transformation in manufacturing. The adoption of digital twins for factories or digital factory twins remain fragmented and often unclear, particularly for small and medium-sized enterprises. This study addresses this ambiguity by systematically deriving archetypes of digital factory twins to support clearer classification, planning, and implementation. Based on a structured literature review and expert interviews, 71 relevant DFT use cases were identified. The result of the conducted cluster analysis is four distinct archetypes: (1) Basic Planning Factory Twin, (2) Advanced Simulation Factory Twin, (3) Integrated Operations Factory Twin, and (4) Holistic Digital Factory Twin. Each archetype is characterized by specific technical features, data integration levels, lifecycle phases, and stakeholder involvement.

Digital Twins (DT) provide detailed, dynamic representations of production systems, but integrating multiple DTs into a distributed ecosystem presents fundamental challenges beyond mere model interoperability. DTs encapsulate dynamic behaviors, optimization goals, and time management constraints, making their coordination a complex, unsolved problem. Moreover, DT development faces broader challenges, including but not limited to data consistency, real-time synchronization, and cross-domain integration, that persist at both individual and distributed scales. This paper systematically reviews these challenges, examines how current research addresses them, and explores their implications in distributed, hierarchical DT environments. Finally, it presents preliminary ideas for a structured approach to orchestrating multiple DTs, laying the groundwork for future research on holistic DT management.

Modeling and Simulation (M&S) has long been essential for decision-making in complex systems due to its ability to explore strategic and operational alternatives in a structured and risk-free manner. The emergence of Digital Twins (DTs) has further enhanced this by enabling real-time bidirectional synchronization with physical systems. However, constructing and maintaining accurate and adaptive models and DTs remains time- and resource-intensive and requires deep domain expertise. In this paper, we introduce an adaptive Decision-Making Framework (DMF) that integrates Large Language Models (LLMs) and Model-Driven Engineering (MDE) into the M&S and DT pipeline. By leveraging LLMs as proxy experts and synthesis aids, and combining them with MDE to improve reliability, our framework reduces manual effort in model construction, validation and decision space exploration. We present our approach and discuss how it improves agility, reduces expert dependency and can act as a pragmatic aid for making enterprise robust, resilient and adaptive.

Modern manufacturing environments, such as semiconductor manufacturing, require agile, data-driven decision support to cope with increasing system complexity. Discrete event simulation (DES) is a key method for evaluating scheduling strategies under uncertainty. Building on a previously introduced low-code framework for scenario farming, this paper presents an extended and evolving approach that integrates machine learning (ML) to enhance scheduling decision support. The framework automates model generation, distributed experimentation, and systematic scenario data collection, providing the basis for training data-driven decision models. Using the Semiconductor Manufacturing Testbed 2020 as a reference, initial experiments demonstrate how simulation-based insights can be transformed into intelligent scheduling aids. Key features include model structure synchronization across simulation tools, automated experiment design, and modular integration of learning components, providing a first step towards adaptive, simulation-driven decision support systems.

In stochastic manufacturing environments, disruptions such as machine breakdowns, variable processing times, and unexpected delays make static scheduling approaches ineffective. To address this, we propose a heuristic-based rolling horizon scheduling method for unrelated parallel machines. The rolling horizon framework addresses system stochasticity by enabling dynamic adaptation through frequent rescheduling of both existing jobs and those arriving within a rolling lookahead window. This method decomposes the global scheduling problem into smaller, more manageable subproblems. Each subproblem is solved using a heuristic approach based on a suitability score that incorporates key factors such as job properties, machine characteristics, and job-machine interactions. Simulation-based experiments show that the proposed method outperforms traditional dispatching rules in dynamic and stochastic manufacturing environments with a fixed number of jobs, achieving shorter makespans and cycle times, reduced WIP levels, and lower machine utilization.

The complexity of job shops is characterized by variable product routing, machine reliability, and operator learning that necessitates intelligent assignment strategies to optimize performance. Traditional models often rely on first-available machine selection, neglecting learning curves and processing time variability. To overcome these limitations, this paper introduces the Data-Driven Job Shop Scheduling (DDJSS) framework, which dynamically selects machines based on the status of resources at the current time steps. To evaluate the effectiveness of the proposed frameworks, we developed two scenarios using FlexSim to perform a thorough analysis. The results demonstrated significant improvements in key performance indicators, including reduced waiting time, lower queue length, and higher throughput. The output is increased by over 144% and 348%, for some exemplary jobs in the case studies mentioned in this paper. This study highlights the value of integrating learning behavior and data-driven assignments for improving decision-making in flexible job shop environments.

This paper investigates a parallel batching problem with incompatible families, unequal job sizes and non-identical machine capacities. The objective is to minimize the total energy costs. Motivated by a real-word autoclave molding process in composite material manufacturing additional factors need to be considered: machine eligibility conditions, machine availability constraints, machine dependent energy consumption, and the job size and machine capacity may not be available in absolute terms. Mathematical models are formulated for both defined and undefined job and machine sizes. The latter approach creates batches based on a set of best-known batches derived from past production data to meet the machine capacity constraint. Finally, a heuristic is presented. Computational experiments are conducted based on a real case study. Energy savings of over 20% can be achieved compared to the actual planning with sensible batch forming and machine allocation in a short amount of computing time.

Rising energy price volatility and the shift toward renewables are driving the need for energy-aware production planning. This paper investigates the integration of energy storage systems into dynamic dispatching to balance production-logistics and energy costs. Building on prior work that introduced a workload- and price-based dispatching rule, we extend the model to include energy storage loading and unloading decisions. The rule prioritizes storage refill at low prices and lets machines resort to stored energy when grid prices rise and the workload is high. A simulation-based evaluation examines scenarios with different storage capacities whereby decision rule parameters are optimized. Computational results demonstrate that a reduction of operational costs is possible without deteriorating production logistics performance. By decoupling energy sourcing from real-time prices, manufacturers achieve both resilience and cost savings. This research contributes to sustainable manufacturing by offering a practical strategy for integrating energy storage into production planning under volatile energy conditions.

In semiconductor manufacturing, Statistical Process Control (SPC) ensures that products meet the Electrical Wafer Sort (EWS) tests performed at the end of the manufacturing flow. In this work, we model the EWS tests for several products using inline SPC data from the Front-End-Of-Line (FEOL) to the Back-End-Of-Line (BEOL). SPC data tend to be inherently sparse because measuring all wafers, lots, and products is both costly and can significantly impact the throughput. In contrast, EWS data is densely collected at the die level, offering high granularity. We propose to model the problem as a regression task to uncover interdependencies between SPC and EWS data at the lot level. By applying two learning strategies, mono- and multi-target, we demonstrate empirically that leveraging families of EWS tests enhances model performance. The performance and practical relevance of the approach are validated through numerical experiments on real-world industrial data.

Cross-process root-cause analysis of wafer defects is among the most critical yet challenging tasks in semiconductor manufacturing due to the heterogeneity and combinatorial nature of processes along the processing route. This paper presents a new framework for wafer defect root cause analysis, called Potential Loss Analysis (PLA), as a significant enhancement of the previously proposed partial trajectory regression approach. The PLA framework attributes observed high wafer defect densities to upstream processes by comparing the best possible outcomes generated by partial processing trajectories. We show that the task of identifying the best possible outcome can be reduced to solving a Bellman equation. Remarkably, the proposed framework can simultaneously solve the prediction problem for defect density as well as the attribution problem for defect scores. We demonstrate the effectiveness of the proposed framework using real wafer history data.

Defect pattern detection and classification present significant challenges in thin-film transistor liquid-crystal display (TFT-LCD) manufacturing. Traditional machine learning approaches struggle with data scarcity, insufficient labeling, and class imbalance. To address these limitations, this study proposes a two-stage defect classification framework for Test for Open/Short (TOS) maps, which measure electrical defects by using contrastive pre-training on semiconductor manufacturing wafer bin maps to enhance classification with limited TOS data, followed by a novel dual-branch open-world semi-supervised learning that robustly handles both class imbalance and novel pattern discovery.

Production planning for wafer fabs often relies on linear programming. Exogenous lead time estimates or workload-dependent lead times by means of clearing functions are taken into account in common planning formulations. For realistic performance assessment purposes, the process uncertainty is captured by simulating the execution of the resulting release schedules, i.e. expected values for profit or costs are considered. In the present paper, we take a more direct approach using simulation-based optimization. The capacity constraints and the lead time representation are indirectly respected by executing release schedules in a simulation model of a large-scaled wafer fab. Variable neighborhood search (VNS) is used to compute release schedules. We show by designed experiments that the proposed approach is able to outperform the allocated clearing function (ACF) formulation for production planning under many experimental conditions.

This paper focuses on a Research and Development (R&D) semiconductor manufacturing system. By virtue of their vocation, R&D facilities tolerate much more variability in processes and outcomes than industrial-scale ones. In such environments, operating under conditions characterized by high uncertainty and occurrences of (un)knowns corresponds to normal operating conditions rather than abnormal ones. This paper characterizes the key entities and operational aspects of a semiconductor R&D cleanroom and introduces a discrete-event simulation model that captures these elements. The simulation model is grounded in empirical data and reflects real-life operations management practices observed in actual R&D cleanroom settings. Preliminary computational results based on real-life instances are presented, and future research directions are outlined to support resilient decision-making in environments where high levels of uncertainty are part of normal operating conditions.

This study investigates which product segments benefit most from dual-fab flexibility in semiconductor manufacturing. Dual-fab flexibility refers to the strategic capability of producing the same product at multiple geographically dispersed fabrication facilities. The focus is on selective flexibility, defined as the targeted allocation of flexibility to specific product segments, such as high-runners (HR: high-volume, frequently ordered products) and low-runners (LR: low-volume, infrequently ordered products). A discrete-event simulation model evaluates four flexibility scenarios under demand uncertainty. Results indicate that prioritizing flexibility for HR products leads to approximately a 7% improvement in cumulative profit values after five years of simulation compared to baseline scenarios. Simulations under 80% forecast accuracy further validate the model’s practical relevance. The findings highlight the value of strategic, demand-driven flexibility in enhancing semiconductor supply chain resilience and provide a foundation for future research on incorporating stochastic demand, dynamic lead times, and advanced forecasting techniques.

Overhead Hoist Transport (OHT) systems significantly influence semiconductor fab performance. Traditional dispatching methods such as Nearest Job First (NJF) focus on minimizing travel distance without considering the dynamic health of vehicles. As a result, unstable vehicles may be dispatched to critical areas, increasing the risk of system-wide disruptions. Previous health-aware approaches often rely on Mean Time To Failure (MTTF), which assumes that failure events are observable—a strong assumption given the uncertainty and latent nature of real-world. To address this limitation, we introduce a Reconstruction Error (RE)-based Health Index (HI) derived from anomaly detection models, enabling dynamic and risk-sensitive dispatching. We define a risk-aware dispatching score that incorporates both traffic exposure and health degradation, and simulate diverse HI profiles with varying predictive fidelity. Our experiments aim to evaluate how the quality of the HI affects dispatching performance and system robustness

This paper addresses the photolithography process scheduling problem, a critical bottleneck in both display and semiconductor production. In display manufacturing, as the number of deposited layers increases and reentrant operations become more frequent, the complexity of scheduling processes has significantly increased. Additionally, growing market demand for diverse product types underscores the critical need for efficient scheduling to enhance operational efficiency and meet due dates. To address these challenges, we propose a novel graph-based reinforcement learning framework that dynamically schedules photolithography operations in real time, explicitly considering mask locations, machine statuses, and associated transfer times. Through numerical experiments, we demonstrate that our method achieves consistent and robust performance across various scenarios, making it a practical solution for real-world manufacturing systems.

We consider job-shop scheduling problems with stress test machines. Several jobs can be processed at the same time on such a machine if the sum of their sizes does not exceed its capacity. Only jobs with operations of the same incompatible family can be processed at the same time on a machine. The machine can be interrupted to start a new or to unload a completed job. A conditioning time is required to reach again the temperature for the stress test. The machine is unavailable during the conditioning process. Operations that cannot be completed before a conditioning activity have to continue with processing after the machine is available again. The makespan is to be minimized. A constraint programming formulation and a variable neighborhood search scheme based on an appropriate disjunctive graph model are designed. Computational experiments based on randomly generated problem instances demonstrate that the algorithms perform well.

The future of leading-edge manufacturing demands innovations that generate real benefits for the business at a pace and scale we have never seen before. This talk will discuss major forces underway shaping manufacturing’s future, transitioning from digital-twins to being increasingly driven by agentic and multi-model solutions that enable full-autonomy and augmented intelligence-based decision support systems in the factory. Their foundations require highly dependable, content-rich, and collaborative environments that can sense, perceive, reason, plan and execute complex production decisions safely around the clock. And as business drivers change, these capabilities must adapt and stay available and resilient, while being extended and scaled out quickly. The talk will outline how these tools are moving from being advisory in nature to the next frontier of execution by objectively characterizing complexity and disruptions autonomously and creating runways that unlock entirely new levels of agility and productivity that manufacturing demands.

This is a panel paper which discusses the use of discrete-event simulation to address problems in semiconductor manufacturing. We have gathered a group of expert semiconductor researchers and practitioners from around the world who (often successfully) applied discrete-event simulation to semiconductor-related problems in the past. The paper collects their answers to an initial set of questions. These serve to not only showcase the current state-of-the-art of discrete-event simulation (DES) in semiconductor manufacturing but also provide insights into where the field is heading and the impact it will have on our world by 2050 in the semiconductor manufacturing domain.

This paper addresses a capacity management problem in semiconductor metrology, where lots are typically sent to measurement under static sampling rules. Such rigid strategies often cause bottlenecks when unexpected events occur, delaying production. To address this, we propose a corrective approach that dynamically selects lots to skip, i.e., not measured, while prioritizing the most critical ones and ensuring that metrology tools respect capacity limits. The method combines a skipping algorithm with the Iterated Min-Max (IMM) workload balancing procedure, which ensures a fair workload distribution and helps identify the most critical tools. Several performance indicators are introduced to evaluate the efficiency of this approach compared to a classical balancing strategy. Computational experiments with industrial data demonstrate that integrating IMM improves lot selection, reduces the number of skipped lots, and preserves measurement for the highest priority ones while better satisfying capacity constraints.

A narrow threshold voltage distribution is essential to ensure uniform performance among parallel-connected power MOSFETs. This work analyzes threshold voltage variation in trench MOSFETs using real data and proposes the use of machine learning models to improve manufacturing uniformity.

The semiconductor industry aims to enhance device performance while maintaining a high yield. Managing variations in process steps is a challenge, traditionally addressed through physical silicon experimentation. This study developed a virtual fabrication model of a FinFET transistor to analyze the impact of process variations, such as etch and deposition steps, on key transistor parameters like threshold voltage (Vth), Drain-Induced Barrier Lowering (DIBL), and subthreshold swing (SS). Analysis of within-wafer variations in spacer nitride thickness revealed significant variability, affecting Vth and yielding 66.7%. Adjusting the nominal spacer nitride deposition thickness and controlling deviations improved yield, confirmed by Monte Carlo simulations. Three process recipes with different deposition thickness distributions were tested, with the "donut" region recipe achieving the highest yield (94.8%) and smallest Vth standard deviation. Additionally, combined variations in spacer nitride deposition and epitaxial SiC growth thickness were analyzed, demonstrating the model's capability to predict and optimize multiple process recipes.

Increasing global demand has led to calls for better methods of process improvement for semiconductor wafer manufacturing. Of these methods, digital twins have emerged as a natural extension of already existing simulation techniques. We argue that despite their extensive use in literature, the current tools used to construct semiconductor simulations are underdeveloped. Without a standardized tool to build these simulations, their modularity and capacity for growth are heavily limited. In this paper, we propose and implement a library of classes in the Python language designed to build on top of the already existing SimPy library. These classes are designed to automatically handle specific common logical features of semiconductor burn-in processes. This design allows users to easily create modular, adaptable, digital twin-ready simulations. Preliminary results demonstrate the library’s efficacy in predicting against benchmark data provided by the Intel Corporation and encourage further development.

This research presents an aggregated simulation model for the front-end semiconductor supply chain to assess master plans, focusing on the impact of demand and supply uncertainties on the key performance indicators on-time delivery and inventory on hand. Supply uncertainty is modeled using discrete distributions of historical cycle times, incorporating load-dependent cycle times through a non-linear regression model. To model demand uncertainty, we use future forecasts and adjust them by sampling from distributions of historical forecast percentage errors. By comparing master plan performance under uncertain conditions with those from deterministic scenarios, the model provides valuable insights into how these uncertainties influence supply chain performance. Using data from NXP Semiconductors N.V., a Dutch semiconductor manufacturing and design company, we demonstrate the model’s applicability and offer practical guidance for industry practitioners. Based on numerical experiments, we conclude that the impact of demand and supply uncertainty significantly differs compared to deterministic planning.

Designing high-fidelity simulation models from natural language descriptions remains a significant challenge, particularly in complex domains like supply chains. Pre-trained large language models (LLMs), when used without domain-specific adaptation, often fail to translate unstructured inputs into executable model representations. In this work, we present a fine-tuned LLM-based pipeline specifically tailored for simulation model generation. We construct a high-quality dataset using a novel data generation process that captures diverse supply chain scenarios. The fine-tuned LLM first converts human-like natural language scenario descriptions into structured representations, then translates these into executable code for a modular Python-based simulation engine built to support a wide range of supply chain configurations. Quantitative and qualitative evaluations show that the fine-tuned model consistently generates high-fidelity simulation models, significantly outperforming pre-trained LLMs in terms of structural accuracy, simulation behavior, and its ability to robustly extract relevant information from linguistically variable natural language descriptions.

Fabless semiconductor companies—semiconductor design experts without their factories—serve as the essential bridge between sophisticated customer needs and technological innovations, playing a pivotal role in the semiconductor supply chain. At these companies, planning teams receive demand forecasts from the sales team and develop production plans that consider inventory, capacity, and lead time. However, due to the inherent characteristics of the semiconductor industry—high demand volatility, short product cycles, and extended lead times—a substantial gap often exists between sales forecasts and actual demand. Consequently, evaluating forecast reliability is critical for planning teams that rely solely on sales forecasts for production planning. In this paper, we propose a novel machine learning framework that assesses forecast reliability by classifying demand forecasts as either overestimates or underestimates rather than using regression methods. Experimental results confirm its effectiveness in assessing forecast reliability.

Accurate demand forecasting is essential for the semiconductor industry to optimize production and meet customer needs. However, challenges like the COVID-19 pandemic, tariffs, and the Bullwhip Effect — the amplification of demand fluctuations along supply chains — create uncertainty and variability. While greater transparency and collaboration could help address these issues, reluctance to share sensitive data across supply chain tiers remains a significant barrier. We present the True Demand Framework for semiconductor supply chains to address these challenges. This framework includes a monthly end-to-end supply chain survey that ensures privacy for all tiers through Multi-Party Computation. The surveys capture key data to reduce variability and connect with annual surveys, linking semiconductor technology nodes to end-market demand. Results are mapped to the semantic Digital Reference model, ensuring interoperability with other surveys and external market data. This enables advanced forecasting and simulation models, improving demand and inventory planning across semiconductor supply chains.

We consider a short-term demand forecasting problem for semiconductor supply chains. In addition to observed demand quantities, order entry information is available. We compute a combinational forecast based on an exponential smoothing technique, a long short-term memory (LSTM) network, and the order entry information. The weights for the different forecast sources and parameters for exponential smoothing are computed using a genetic algorithm. Computational experiments based on a rich data set from a semiconductor manufacturer are conducted. The results demonstrate that the best forecast performance is obtained if all the different forecasts are combined.

The complexity of modern semiconductor fabrication (FAB) systems makes it difficult to implement integrated simulation systems that combine production and logistics simulators. As a result, these simulators have traditionally been developed independently. However, in actual FAB operations, information exchange between Real-Time Schedulers (RTS) and Real-Time Dispatchers (RTD) coordinates production activities. To address this issue, we propose a coupled RTS–RTD simulation framework that integrates production and logistics simulators into a unified environment. In addition, we introduce a dynamic decision-making rule that enables flexible responses when logistical constraints prevent execution of the original production schedule. Simulation experiments were conducted using the SMT2020 and SMAT2022 datasets. The results show that selectively following RTD decisions, instead of strictly adhering to RTS-generated schedules, can significantly improve production efficiency in FAB operations.

We consider a scheduling problem for a two-stage flexible flow shop with maximal time lags between consecutive operations motivated by semiconductor manufacturing. The jobs have unequal ready times and both initial and inter-stage time lags. The total weighted tardiness is the performance measure of interest. A heuristic scheduling framework using genetic programming (GP) to automatically discover priority indices is applied to the scheduling problem at hand. Computational experiments are carried out on randomly generated problem instances. The results are compared with the ones of a reference heuristic based on a biased random-key genetic algorithm combined with a backtracking procedure and a mixed-integer linear programming-based decomposition approach. The results show that high-quality schedules are obtained in a short amount of computing time using the GP approach.

Semiconductor fabs must expedite urgent wafer carriers, known as Hot Lots, yet excessive priority can impair throughput. DAIM Research Corp. categorizes priority into three levels; we address Level 2, where the Hot Lot OHTs wins all local contests for path selection and merging while other moves proceed. We present a two-layer control framework. First, extending the reinforcement-learning-based dynamic routing algorithm, temporarily modify the expected travel time of tracks ahead the Hot Lot, causing regular OHTs to detour proactively. Second, modifying the entrance sequence at the merge of two tracks to reduce Hot Lot stops and waits. The effectiveness of the framework is demonstrated through simulation-based experiments, which show that Hot Lot travel time is shortened with acceptable impact on overall throughput.

Accurate cycle time prediction is critical for optimizing throughput, managing WIP, and ensuring responsiveness in semiconductor manufacturing. This systematic review synthesizes literature from Google Scholar, Web of Science, and IEEE Xplore, covering analytical, statistical, AI-driven, and hybrid approaches. The key contributions are: (1) a structured, comparative evaluation of predictive techniques in terms of accuracy, interpretability, and scalability, and (2) identification of research gaps and emerging directions, such as self-adaptive models, generative AI for data augmentation, and enhanced human-AI collaboration. This review provides insights to support the development of robust forecasting systems aligned with the evolving demands of semiconductor manufacturing.

In today's volatile and complex world, non-manufacturing processes are crucial for global agile and, thus, resilient semiconductor supply chains. Examples of non-manufacturing processes include new product samples, fab overarching process flows, and enabling processes. This conceptual paper begins with the hypothesis that operating curve principles from manufacturing can also be applied to non-manufacturing processes. To achieve this, we needed to understand why the principles of operating curves and flow factors lead to efficiency and effectiveness in semiconductor manufacturing. Our goal was to determine how these principles can be applied to broader contexts and specific use cases. The initial experimental results are promising. Furthermore, it is essential to assess how non-manufacturing processes are structured. The paper concludes with examples demonstrating the potential to bridge the efficiency gap between manufacturing and non-manufacturing processes. The ultimate goal is to match both efficiency levels, thus opening new opportunities within global semiconductor supply chains.

This research introduces a dynamic risk control framework for a two-stage production system with downstream queue time constraints. Queue time constraints are critical in semiconductor manufacturing, as their violations lead to quality degradation and production losses. To reduce the risk of violating queue time constraints, a key principle in queue time constraint management is controlling upstream production when downstream queue times exceed critical thresholds. Traditional control methods face challenges in accurately responding to dynamic manufacturing conditions. Our approach develops a dynamic admission control method that predicts queue times by estimating the real-time downstream processing capacity and the number of preceding jobs. We implement a two-stage system combining upstream admission control with downstream priority dispatching. Through simulation experiments in multi-machine, multi-product systems, we evaluate performance across various utilization rates and capacity configurations. Results show that our approach reduces total costs, including scrap and inventory holding costs, while maintaining optimal throughput.

Ontologies are essential for structuring knowledge in complex domains like semiconductor supply chains but often remain inaccessible to non-technical users. This paper introduces a combined classical and AI-based approach to improve ontology explainability, using Digital Reference (DR) as a case study. The first approach leverages classical ontology visualization tools, enabling interactive access and feedback for user engagement. The second integrates Neo4j graph databases and Python with a large language model (LLM)-based architecture, facilitating natural language querying of ontologies. A post-processing layer ensures reliable and accurate responses through query syntax validation, ontology schema verification, fallback templates, and entity filtering. The approach is evaluated with natural language queries, demonstrating enhanced usability, robustness, and adaptability. By bridging the gap between traditional query methods and AI-driven interfaces, this work promotes the broader adoption of ontology-driven systems in the Semantic Web and industrial applications, including semiconductor supply chains.

A key concern for factory operators is the prioritization of machine groups for preventive maintenance to reduce downtimes of the most critical machine groups. Typically, such maintenance decisions are made based on heuristic rules supported by discrete-event simulations. Here, we present a large-scale sensitivity analysis that elucidates the relationship between machine group features and factory-level performance indicators in the presence of unplanned tool downtimes in SMT2020. From extensive simulation data, we extract predictors that permit a machine group ranking according to their criticality. An exhaustive combinatorial evaluation of 11 machine group features allows us to identify the most predictive features at different subset sizes. By evaluating and visualizing predictions made using linear regression models and neural networks against the ground truth data, we observe that to accurately rank machine groups, capturing nonlinearities is vastly more important than the size of the feature set.

The semiconductor industry faces complex challenges in scheduling, demanding efficient solutions. Reinforcement Learning (RL) holds significant promise; however, transitioning it from research to a productized solution requires overcoming key challenges such as reliability, tunability, scalability, delayed rewards, ease of use for operators, and connectivity with legacy systems. This paper explores these challenges and proposes strategies for effective productization of RL in wafer scheduling as developed by Applied Materials’ AI/ML team in the Applied SmartFactory® offering. In the context of this paper, “productization” refers specifically to the engineering and system-level integration needed to make RL solutions operational in fabs.

In this study, we propose a two-stage stochastic programming method to improve plan stability in semiconductor supply chain master planning in a rolling horizon setting. The two-stage programming model is applied to real-world data from NXP Semiconductors N.V. to assess the quality of generated plans based on the KPIs plan stability, on-time delivery, and inventory position. We also compare the performance of two-stage stochastic programming to linear programming. To model demand uncertainty, we propose to fit distributions to historical demand data from which stochastic demand can be sampled. For modeling supply, we propose an aggregated rolling horizon simulation model of the front-end supply chain. Based on the performed experiments, we conclude that two-stage programming outperforms LP in terms of plan stability, while performing comparably in terms of inventory position and on-time delivery.

Complex supply networks, volatile demand, and up to 26-week lead times pose severe trade-offs between inventory efficiency and service flexibility in semiconductor supply chains. We develop a discrete-event simulation to compare Order-up-to policies against an Available-to-promise (ATP) approach. Results show that ATP dynamically reallocates capacity, cuts reliance on finished-goods buffers, and boosts operational agility without eroding financial performance. Our findings suggest integrating ATP with real-time decision support can reconcile efficiency and flexibility under extreme lead-time uncertainty by enabling proactive order adjustments, decreasing backorders, and cost overruns.

In semiconductor wafer planning, long lead times and volatile demand signals pose significant challenges. Reacting to every demand change introduces nervousness and inefficiencies, particularly in backend operations. This case study presents NXP's Wafer Workbench, a decision-support tool that enables planners to assess demand changes through simulation under various capacity scenarios. By comparing new demand signals with existing wafer plans using KPIs such as Expected Service Level and Capacity Utilization, the tool enhances visibility and coordination. The Wafer Workbench has significantly reduced planning workload and infeasibilities while improving plan quality. Future enhancements include flexible scenario simulation, stochastic optimization to address demand uncertainty, and generative AI agents for automated plan adjustments.

Efficient crowd control in public spaces is critical for mitigating threats and ensuring public safety, especially in scenarios where live testing environments are limited. It is important to study crowd behavior following disruptions and strategically allocate law enforcement resources to minimize the impact on civilian populations to improve security systems and public safety. This paper proposes an extended social force model to simulate crowd evacuation behaviors in response to security threats, incorporating the influence and coordination of law enforcement personnel. This research examines evacuation strategies that balance public safety and operational efficiency by extending social force models to account for dynamic law enforcement interventions. The proposed model is validated through physics-based simulations, offering insights into effective and scalable solutions for crowd control at public events. The proposed hybrid simulation model explores the utility of integrating agent-based and physics-based approaches to enhance community resilience through improved planning and resource allocation.

Timely and accurate prediction of adversarial unit movements is a critical capability in military operations, yet traditional methods often lack the granularity or adaptability to deal with sparse, uncertain data. This paper presents a probabilistic isochrone (PI) framework to estimate future positions of military units based on sparse reconnaissance reports. The approach constructs a continuous probability density function of movement distances and derives gradient prediction areas. Validation is conducted using real-world data from the 2022 Russian invasion of Ukraine, evaluating both the inclusion of actual future positions within the predicted rings and the root mean-squared error of our method. Results show that the method yields reliable spatial uncertainty bounds and offers interpretable predictive insights. This PI approach complements existing isochrone mapping and adversarial modeling systems and demonstrates a novel fusion of simulation, spatial analytics, and uncertainty quantification in military decision support. Future work will integrate simulation to enhance predictive fidelity.

Realistic crowd modeling is essential for military and security simulation models. In this paper we address modeling of the movement of people in the types of unstructured crowds that are common in civil security situations. Early approaches in the literature to simulating the movement of individuals in a crowd, typically treated the crowd as consisting of entities moving on a fixed grid, or as particles in a fluid flow, where the movement rules were relatively simple and each member had the same goal, such as to move along a crowded sidewalk or to evacuate through an exit. This paper proposes a 2-part approach for more complex pedestrian movement modeling that takes into account the cognitively-determined behavioral intent of each member of the crowd to determine their own movement objective while also allowing each to temporarily react to a short-term urgent situation that may arise while pursuing their movement goal.

NATO allies are preparing for Large-Scale Combat Operations (LSCOs) against peer or near-peer adversaries. Although a significant increase in casualties with life-threatening injuries is expected, western military personnel lack experience with the medical requirements of LSCOs. We propose the use of simulation to conduct necessary research, estimate the resources required, and adapt the doctrine. We therefore present a scenario for assessing NATO’s clinical timelines based on open-access data, showing that a shortage of surgical capacity is likely to occur.

Aircraft fleet transitions are about more than obtaining a new platform. The transition of personnel, support, and operational responsibility makes transitions complex and fraught with risk. Here we discuss Defence Research and Development Canada (DRDC) Defence Scientists’ approach to evaluating Royal Canadian Air Force (RCAF) fleet transitions. We delve into the use of two different toolsets to model transitions.

A scenario of casualty evacuations from the frontlines in Ukraine was simulated in SIMEDIS, incorporating persistent drone threats that restricted daytime evacuations. A stochastic discrete-event approach modeled casualty location and health progression. Casualties from a First-Person View drone explosion in a trench were simulated, incorporating controlled versus uncontrolled bleeding in rescue and stabilization efforts. Two evacuation strategies were compared: (A) transport to a nearby underground hospital with delays and (B) direct transport to a large hospital with potential targeting en route. Results showed that strategy A was safer for transport, but effective hemorrhage control was crucial for survival. Strategy A led to lower mortality than strategy B only when hemorrhage control was sufficient. Without it, both strategies resulted in similar mortality, emphasizing that blood loss was the primary cause of death in this simulation.

This paper introduces Echo Warfare, a novel non-kinetic doctrine that strategically targets intangible cultural heritage through systematic replication, appropriation, and institutional rebranding. Unlike traditional psychological or cognitive warfare, Echo Warfare aims for the intergenerational erosion of cultural identity and memory infrastructure. While current international legal frameworks inadequately address such systematic cultural appropriation, this paper advances discourse by implementing a Dynamic Bayesian Network (DBN) simulation to model how cultural anchors degrade under sustained echo feedback. The simulation reveals threshold effects, legitimacy dynamics, and population-level vulnerabilities, offering a predictive modeling tool for identifying intervention thresholds and patterns of vulnerability. The study concludes with legal, educational, and simulation-based policy recommendations to mitigate the long-term effects of Echo Warfare.

This study presents an agent-based simulation that examines how pre-attack misinformation amplifies the effectiveness of spearphishing campaigns within organizations. A virtual organization of 235 end user agents is modeled, each assigned unique human factors such as Big Five personality traits, fatigue, and job performance, derived from empirical data. Misinformation is disseminated through Facebook, where agents determine whether to believe and spread false content using regression models from prior psychological studies. When agents believe misinformation, their psychological and organizational states degrade to simulate a worst-case scenario. These changes increase susceptibility to phishing emails by impairing security-related decision-making. Informal relationship networks are constructed based on extraversion scores, and network density is varied to analyze its effect on misinformation spread. The results demonstrate that misinformation significantly amplifies organizational vulnerability by weakening individual and collective cybersecurity-relevant decision-making, emphasizing the critical need to account for human cognitive factors in future cybersecurity strategies.

Disinformation campaigns can distort public perception and destabilize institutions. Understanding how different populations respond to information is crucial for designing effective interventions, yet real-world experimentation is impractical and ethically challenging. To address this, we develop an agent-based simulation using Large Language Models (LLMs) to model responses to misinformation. We construct agent personas spanning five professions and three mental schemas, and evaluate their reactions to news headlines. Our findings show that LLM-generated agents align closely with ground-truth labels and human predictions, supporting their use as proxies for studying information responses. We also find that mental schemas, more than professional background, influence how agents interpret misinformation. This work provides a validation of LLMs to be used as agents in an agent-based model of an information network for analyzing trust, polarization, and susceptibility to deceptive content in complex social systems.

The growing complexity of modern warfare demands advanced AI-driven decision support for validating Operational Plans (OPLANs). This paper proposes a multi-agent reinforcement learning framework integrated into the ReLeGSim environment to rigorously test military strategies under dynamic conditions. The adoption of deep reinforcement learning enables agents to learn optimal behavior within operational plans, transforming them into “intelligent executors”. By observing these agents, one can identify vulnerabilities within plans. Key innovations include: (1) a hybrid approach combining action masking for strict OPLAN adherence with interleaved behavior cloning to embed military doctrine; (2) a sequential training approach where agents first learn baseline tactics before evaluating predefined plans; and (3) data farming techniques using heatmaps and key performance indicators to visualize strategic weaknesses. Experiments show hard action masking outperforms reward shaping for constraint enforcement. This work advances scalable, robust AI-driven OPLAN validation through effective domain knowledge integration.

This paper explores the application of deep learning and big data analytics to assess Weapon Combat Effectiveness (WCE) in dynamic combat scenarios. Traditional WCE models rely on simplified assumptions and limited input variables, limiting their realism. To overcome these challenges, datasets are generated from integrated Virtual-Constructive (VC) simulation frameworks, combining strengths of defense modeling, big data, and artificial intelligence. A case study features two opposing forces: a Blue Force (seven F-16 aircraft) and a Red Force (two surface-to-air missile (SAM) units) and a high-value facility. Raw simulation data is processed to extract Measures of Performance (MOPs) to train a convolutional neural network (CNN), to capture nonlinear relationships and estimate mission success probabilities. Results show the model’s resilience to data noise and its usefulness in generating decision-support tools like probability maps. Early results suggest that deep learning integrated with federated VC simulations can significantly enhance fidelity and flexibility of WCE analytics.

Supply chain resiliency analysis has become a critical focus for organizations striving to navigate disruptions and uncertainties in a complex, global economy. For products requiring a multitude of components and suppliers, the magnitude of the impact of disruptions increases exponentially. Anticipating and addressing these potential disruptions necessitates advanced modeling approaches that provide actionable insights into how supply chain dynamics unfold under both normal and disrupted conditions. Researchers at Pacific Northwest National Laboratory (PNNL) developed a supply chain resiliency model using Process Simulator, a discrete event simulation (DES) software, to model the fabrication of assemblies. This DES evaluates the performance of an assembly’s supply chain as a whole and its ability to keep pace with demand given the individual supply chains of each component used in the fabrication of the assemblies.

The cyber vulnerability terrain is largely amplified in critical infrastructure systems (CISs) that attract exploitative (nation-state) adversaries. This terrain is layered over an IT and IoT-driven operational technology (OT) network that supports CIS software applications and underlying protocol communications. Usually, the network is too large for both cyber adversaries and defenders to control every network resource under budget constraints. Hence, both sides strategically want to target 'crown jewels' (i.e., critical network resources) as points of control in the IT/OT network. Going against traditional CIS game theory literature that idealistically (impractically) model attacker-defense interactions, we are the first to formally model real-world adversary-defender strategic interactions in CIS networks as a simultaneous non-cooperative network game with an auction contest success function (CSF) to derive the optimal defender strategy at Nash equilibria. We compare theoretical game insights with those from large-scale Monte Carlo game simulations and propose CIS-managerial cyber defense action items.

As cyber threats become more sophisticated, rapid and accurate vulnerability detection is essential for maintaining secure systems. This study explores the use of Large Language Models in software vulnerability assessment by simulating the identification of Python code with known Common Weakness Enumerations (CWEs), comparing zero-shot, few-shot cross-domain, and few-shot in-domain prompting strategies. Our results indicate that few-shot prompting significantly enhances classification performance, particularly when integrated with confidence-based routing strategies that improve efficiency by directing human experts to cases where model uncertainty is high.

We find that LLMs can effectively generalize across vulnerability categories with minimal examples, suggesting their potential as scalable, adaptable cybersecurity tools in simulated environments. By integrating AI-driven approaches with expert-in-the-loop (EITL) decision-making, this work highlights a pathway toward more efficient and responsive cybersecurity workflows. Our findings provide a foundation for deploying AI-assisted vulnerability detection systems that enhance resilience while reducing the burden on human analysts.

Modern modelling and simulation techniques allow us to safely test the policies used to mitigate disasters. We show how the DEVS formalism can be used to ease the modelling process by exploiting its modularity. We show how a policymaker’s existing models of any type can be recreated with DEVS so they may be reused in any new models, decreasing the number of new models that need to be made. We recreate a sequential decision model of an arctic major maritime disaster developed by the Canadian government as a DEVS model to demonstrate the method. The case study shows how DEVS allows policymakers to create models for studying emergency policies with greater ease. This work shows a method that can be used by policymakers, including models of emergency scenarios, and how they can benefit from creating equivalent DEVS models, as well as exploiting the beneficial properties of the DEVS formalism.

Wildfires are extremely dangerous and destructive. In order to protect populations and infrastructure, government officials and firefighters must best decide how to expend their limited resources. Wildfire simulation aids these decision makers by predicting the spread of fire. One method to simulate wildfires is using Cellular Automata. In this paper we present a hybrid model combining a cellular wildfire spread model using asymmetric Cell-DEVS models, the Behave library for calculating the rate spread of the fire (based on difference equations), and a Geographical Information System. The GIS information from publicly available maps, and the combination of techniques can improve the accuracy of the wildfire spread predictions, and allows multiple scenarios to be simulated in timely manner.

Behavioral models of component-based dynamical systems are integral to building useful simulations. Toward this goal, approaches enabled by Large Language Models (LLMs) have been proposed and developed to generate grammar-based models for Discrete Event System Specification (DEVS). This paper introduces PDEVS-LLM, an agentic framework to assist in developing Parallel DEVS (PDEVS) models. It proposes using LLMs with statecharts to generate behaviors for parallel atomic models. Enabled with PDEVS concepts, plausible facts from the whole description of a system are extracted. The PDEVS-LLM is equipped with grammars for the PDEVS statecharts and hierarchical coupled model. LLM agents assist modelers in (re-)generating atomic models with conversation histories. Examples are developed to demonstrate the capabilities and limitations of LLMs for generative PDEVS models.

Agent-based simulations for networked anagram games, often taking advantage of the experimental data, are useful tools to investigate collaborative behaviors. To confidently incorporate the statistical analysis from the experimental data into the ABS, it is crucial to conduct sufficient validation for the underlying statistical models. In this work, we propose a systematic approach to evaluate the validity of statistical methods of players’ action sequence modeling for the networked anagram experiments. The proposed method can appropriately quantify the effect and validity of expert-defined covariates for modeling the players’ action sequence data. We further develop a Large Language Model (LLM)-guided method to augment the covariate set, employing iterative text summarization to overcome token limits. The performance of the proposed methods is evaluated under different metrics tailored for imbalanced data in networked anagram experiments. The results highlight the potential of LLM-driven feature discovery to refine the underlying statistical models used in agent-based simulations.

The correct design of digital value stream models is an intricate task, which can be challenging especially for untrained or inexperienced users. We address the question whether large language models can be adapted to "understand" value stream’s structure and act as modeling assistants, which could support users with repairing errors and adding or configuring process steps in order to create valid value stream maps that can be simulated. Specifically, we propose a domain-specific multi-task training process, in which an instruction-tuned large language model is fine-tuned to yield specific information on its input value stream or to fix scripted modeling errors. The resulting model – which we coin Llama-VaStNet – can manipulate value stream structures given user requests in natural language. We demonstrate experimentally that Llama-VaStNet outperforms its domain-agnostic vanilla counterpart, i.e. it is 19% more likely to produce correct individual manipulations.

Mission Engineering (ME) requires coordination of multiple systems and stakeholders, but often suffers from unclear problem definitions, fragmented knowledge, and limited engagement. This paper proposes a hybrid methodology integrating Retrieval-Augmented Generation (RAG), Human-Centered Design (HCD), and Participatory Design (PD) within a Model-Based Systems Engineering (MBSE) framework. The approach generates context-rich, stakeholder-aligned mission problem statements, as demonstrated in the Spectrum Lab case study, ultimately improving mission effectiveness and stakeholder collaboration.

This paper employs System Dynamics (SD) to model and analyze DC power distribution systems, focusing on methodological development and using microgrids as case studies. The approach follows a bottom-up methodology, starting with the fundamentals of DC systems and building toward more complex configurations. We coin this approach “Power Dynamics,” which uses stocks and flows to represent electrical components such as resistors, batteries, and power converters. SD offers a time-based, feedback-driven approach that captures component behaviors and system-wide interactions. This framework provides computational efficiency, adaptability, and visualization, enabling the integration of control logic and qualitative decision-making elements. Three case studies of microgrids powered by renewable energy demonstrate the framework’s effectiveness in simulating energy distribution, load balancing, and dynamic power flow. The results highlight SD’s potential as a valuable modeling tool for studying modern energy systems, supporting the design of flexible and resilient infrastructures.

This paper proposes an eigenvalue-based small-sample approximation of the celebrated Markov Chain Monte Carlo that delivers an invariant steady-state distribution that is consistent with traditional Monte Carlo methods. The proposed eigenvalue-based methodology reduces the number of paths required for Monte Carlo from as many as 1,000,000 to as few as 10 (depending on the simulation time horizon T), and delivers comparable, distributionally robust results, as measured by the Wasserstein distance. The proposed methodology also produces a significant variance reduction in the steady-state distribution.

Human decision-making under uncertainty poses a significant challenge in modeling disrupted pharmaceutical supply chains. This study introduces a behaviorally grounded simulation framework systematically integrating empirical data from participatory game experiments using an advanced methodological pipeline. Specifically, we combine Principal Component Analysis (PCA) for dimensionality reduction, Longest Common Prefix (LCP) clustering for behavioral sequence segmentation, and Hidden Markov Models (HMMs) for dynamic behavioral archetype extraction. These integrated methods identify three archetypes: Hoarders, Reactors, and Followers, which we embed into agent-based simulations to assess impacts under varying disruption and information-sharing scenarios. Agent realism is rigorously validated using Kolmogorov-Smirnov tests that compare simulated and empirical behavioral data, demonstrating strong alignment in general but revealing notable discrepancies among Hoarders under disruption-duration conditions. The findings highlight the methodological necessity of capturing dynamic behavioral heterogeneity, guiding future research on alternative clustering methods, multiphase clustering, and dynamic behavioral transition modeling for enhanced supply chain resilience.

Utilizing data-dependence analysis (DDA) in parallel discrete-event simulation (PDES) to find event-level parallelism, we present the DDA-PDES framework as an alternative to spatial-decomposition (SD) PDES. DDA-PDES uses a pre-computed Independence Time Limit (ITL) table to efficiently identify events in the pending-event set that are ready for execution, in a shared-memory-parallel simulation engine. Experiments with AMD, Qualcomm, and Intel platforms using several packet-routing network models and a PHOLD benchmark model demonstrate speedup of up to 8.82x and parallel efficiency of up to 0.91. In contrast with DDA-PDES, experiments with similar network models in ROSS demonstrate that SD-PDES cannot speed up the packet-routing models without degradation to routing efficacy. Our results suggest DDA-PDES is an effective method for parallelizing discrete-event simulation models that are computationally intensive, and may be superior to traditional PDES methods for spatially-decomposed models with challenging communication requirements.

Parallel Discrete Event Simulation (PDES) enables model developers to run massive simulations across millions of computer nodes. Unfortunately, writing PDES models is hard, especially when implementing reverse handlers. Reverse handlers address the issue of saving the entire state of Logical Processes (LPs), a necessary step for optimistic simulation. They restore the state of LPs through reverse computation, which reduces memory and computation for most common models. This forces model designers to implement two tightly coupled functions: an event handler and a reverse handler. A small bug in a reverse handler can trigger a cascade of errors, ultimately leading to the simulation’s failure. We have implemented a strategy to check reverse handlers one event at a time. In a mature PDES model, we identified reverse handling bugs that would have required processing approximately 1 billion events to manifest, which is nearly impossible to detect using traditional debugging techniques.

warped2 is a general purpose discrete event simulation kernel that contains a robust event time comparison mechanism to support a broad range of modeling domains. The warped2 kernel can be configured for sequential, parallel, or distributed execution. The parallel or distributed versions implement the Time Warp mechanism (with its rollback and relaxed causality) such that a total order on events can be maintained. To maintain a total order, warped2 has an event ordering mechanism that contains up to 10 comparison operations. While not all comparisons require evaluation of all 10 relations, the overall cost of time comparisons in a warped2 simulation can still consume approximately 15-20% of the total runtime. This work examines the runtime costs of time comparisons in a parallel configuration of the warped2 simulation kernel. Optimizations to the time comparison mechanism are explored and the performance impacts of each are reported.

Validation and documentation of rationale are central to simulation studies. Most current approaches focus only on individual simulation artifacts---most typically simulation models---and their validity rather than their contribution to the overall simulation study. Approaches that aim to validate simulation studies as a whole either impose structured processes with the implicit assumption that this will ensure validity, or they rely on capturing provenance and rationale, most commonly in natural language, following accepted documentation guidelines. Inspired by dialectic approaches for developing mathematical proofs, we explore the feasibility of capturing validity and rationale information as a study unfolds through agent dialogs that also generate the overall simulation-study argument. We introduce a formal framework, an initial catalog of possible interactions, and a proof-of-concept tool to capture such information about a simulation study. We illustrate the ideas in the context of a cell biological simulation study.

Automatically generating and executing simulation experiments promises to make running simulation studies more efficient, less error-prone, and easier to document and replicate. However, during experiment generation, background knowledge is required regarding which experiments using which inputs and outputs are useful to the modeler. Therefore, we conducted an interview study to identify what types of experiments modelers perform during simulation studies. From the interview results, we defined four general goals for simulation experiments: exploration, confirmation, answering the research question, and presentation. Based on the goals, we outline and demonstrate an approach for automatically generating experiments by utilizing an explicit and thoroughly detailed conceptual model.

With the rapid advancement of computing technology, there has been growing interest in effectively solving large-scale ranking and selection (R&S) problems. In this paper, we propose a new large-scale fixed-budget R&S procedure, namely the enhanced upper confidence bound (EUCB) procedure. The EUCB procedure incorporates variance information into the dynamic allocation of simulation budgets. It selects the alternative with the largest upper confidence bound. We prove that the EUCB procedure has sample optimality; that is, to achieve an asymptotically nonzero probability of correct selection (PCS), the total sample size required grows at the linear order with respect to the number of alternatives. We demonstrate the effectiveness of the EUCB procedure in numerical examples. In addition to achieving sample optimality under the PCS criterion, our numerical experiments also show that the EUCB procedure maintains sample optimality under the expected opportunity cost (EOC) criterion.

Conventional population-based ODE models struggle against increased level of resolution since incorporating many states exponentially increases computational costs, and demands robust calibration for numerous hyperparameters. PySIRTEM is a spatiotemporal SEIR-based epidemic simulation platform that provides high resolution analysis of viral disease progression and mitigation. Based on the authors-developed Matlab© simulator SIRTEM, PySIRTEM’s modular design reflects key health processes, including infection, testing, immunity, and hospitalization, enabling flexible manipulation of transition rates. Unlike SIRTEM, PySIRTEM uses a Sequential Monte Carlo (SMC) particle filter to dynamically learn epidemiological parameters using historical COVID-19 data from several U.S. states. The improved accuracy (by orders of magnitude) make PySIRTEM ideal for informed decision-making by detecting outbreaks and fluctuations. We further demonstrate PySIRTEM ’s usability performing a factorial analysis to assess the impact of different hyperparameter configurations on the predicted epidemic dynamics. Finally, we analyze containment scenarios with varying trends, showcasing PySIRTEM ’s adaptability and effectiveness.

We introduce a novel iheap module, a flexible Python library for heap operations. The iheap module introduces a generalized comparator function, making it more flexible and general compared to the heapq that is commonly used. We demonstrate that the iheap module achieves parity in terms of time complexity and memory usage against established standard heap modules in Python. Furthermore, the iheap module provides advanced methods and customization options unavailable in its counterparts, which enable the user to implement the heap operations with greater flexibility and control. We demonstrate the iheap module through two case studies. The first case study focuses on the efficient allocation of funds across marketing campaigns with uncertain returns. The second case study focuses on patient triaging and scheduling in the emergency rooms of hospitals. The iheap module provides a powerful and easy-to-use tool for heap operations commonly used in simulation studies.

Neglecting the dangers of decommissioning offshore gas and oil facilities can lead to significant environmental destruction, safety risks, legal issues, and other costs that are likely to increase as these outdated structures deteriorate. This paper provides a comprehensive review of risk modeling techniques employed in offshore oil and gas decommissioning projects. Through systematic literature analysis, relevant papers published between 2015 and 2025, the key risk modeling approaches: multi-criteria decision-making frameworks, Bayesian network applications, and sensor/digital-twin/physics-based modeling techniques, were identified. The paper evaluates each type of method's advantages, limitations, and application scope under different decommissioning scenarios. Despite significant advances in risk modeling, persistent gaps exist due to the uncertainty and complexity of the projects. This review serves as a foundation for understanding the current state of risk modeling in offshore decommissioning while identifying critical areas for future research development.

The need for standardized exchange of dynamic models led to the Functional Mockup Interface (FMI), which facilitates model exchange and co-simulation across multiple tools. Integration of this standard with modeling and simulation formalism enhances interoperability and provides opportunities for collaboration. This research presents an approach for the integration of FMI and Discrete Event System Specification (DEVS). DEVS provides the modularity required for seamlessly integrating the shared model. We proposed a framework for exporting and co-simulating DEVS models as well as for importing and co-simulating continuous-time models using the FMI standard. We present a case study that shows the use of this framework to simulate the steering system of an Unmanned Ground Vehicle (UGV).

A simulation conceptual model (SCM) has many uses such as supporting clarification of the model design across stakeholders, guiding the computer model development, and supporting verification and validation efforts; however, proper documentation of the SCM is critical. SCM literature covers approaches and methods for developing SCMs, and for documenting SCMs, but literature directly addressing content considerations for SCM documentation is sparse. In this paper we address the determination of what content should be include in SCM documentation to maximize its utility. We first synthesize information across the literature, creating a superset of types of SCM content, and also discuss different audiences and their purposes for SCM documentation. We then provide a first-pass analysis of what types of content are most important to include in SCM documentation for different audiences and purposes, offering insights for determining which types of content to include in any given model's documentation.

This article provides an integrated epistemological and ontological frame of reference for scholars and practitioners in the field of gaming and simulation. This frame lies the foundation for an advancement in the field by developing a definition of game science. The article defines what games are (artifacts consisting of rules, roles, and resources), how they contribute to knowledge creation (‘knowing’ as doing) and how they connect the analytical and design sciences. Gaming and simulation are means to understand and design complex realities and therefore are an important means for actors to deal with the complexities of our world. However, we experience a stagnation of the science level in our discipline in a fast-moving world. To enable future and emerging gaming and simulation researchers to practice game science in a rigorous way, we have formulated statements and used an existing science framework to define game science as discipline.

This paper presents a comparative study of AI-based decision agents for real-time control in construction manufacturing, a field traditionally ill-suited to mass production due to high customization and uncertainty. Three approaches are developed and evaluated: an experience-based agent trained through reinforcement learning (EDA), a search-based agent using local stochastic sampling (SDA), and a hybrid combining both strategies (HybridDA). The goal is to minimize component delivery delays. A simulation model of a real precast concrete factory helps generate training patterns and assess production performance. Results show that the EDA significantly outperforms the SDA, despite relying on generalized experience rather than task-specific knowledge. The HybridDA provides modest gains, especially as search effort increases. These findings highlight the unexpected effectiveness of experience-based agents in dynamic manufacturing environments and the potential of hybrid approaches. Future work will explore alternative learning strategies, extend the scope of manufacturing decisions, and evaluate latency performance across the agents.

Autonomous drones play a critical role in dynamic environments, with effective decision making hinging on timely and accurate belief updating. However, the mechanisms underlying AI belief updating, especially under uncertainty and partial observability, remain unclear. This study introduces a novel framework combining Bayesian inference, evidence accumulation theory, and Dynamic Time Warping (DTW) to analyze the AI belief updating process. Experimental validation conducted using simulated urban search and rescue (USAR) scenarios reveals that drone decisions robustly prioritize environmental factors like wind conditions. Moreover, the duration of belief updates inversely relates to environmental change intensity, showcasing adaptive cognition like humans. A dual threshold system that captures belief updates independent of behavior changes, differentiating internal cognitive shifts from observable behavioral actions, thereby enhancing alignment with human cognitive patterns. These findings offer foundational insights contributing toward transparent, adaptive, and human-aligned AI.

The development of object detection models for construction safety is often limited by the availability of high-quality, annotated datasets. This study explores the use of synthetic images generated by DALL·E 3 to supplement or partially replace real data in training YOLOv8 for detecting construction-related objects. We compare three dataset configurations: real-only, synthetic-only, and a mixed set of real and synthetic images. Experimental results show that the mixed dataset consistently outperforms the other two across all evaluation metrics, including precision, recall, IoU, and mAP@0.5. Notably, detection performance for occluded or ambiguous objects such as safety helmets and vests improves with synthetic data augmentation. While the synthetic-only model shows reasonable accuracy, domain differences limit its effectiveness when used alone. These findings suggest that high-quality synthetic data can reduce reliance on real-world data and enhance model generalization, offering a scalable approach for improving construction site safety monitoring systems.

This paper aims to derive a simulation model to evaluate the impact of bus fleet electrification on service reliability. At its core, the model features a micro discrete event simulation (DES) of an urban bus network, integrating a route-level bus operation module and a stop-level passenger travel behavior module. Key reliability indicators—bus headway deviation ratio, excess passenger waiting time, and abandonment rate—are computed to assess how varying levels of electrification influence service reliability. A case study of route 35 operated by DASH in Alexandria, VA, USA is conducted to demonstrate the applicability and interpretability of the developed DES model. The results reveal trade-offs between bus fleet electrification and service reliability, highlighting the role of operational constraints and characteristics of electric buses (EBs). This research provides transit agencies with a data-driven tool for evaluating electrification strategies while maintaining reliable and passenger-centered service.

Flood risk communication influences how the public perceives hazards and motivates preparedness actions, like taking preventive measures, purchasing insurance, and making informed property decisions. Prior research suggests that both the format and source of information can influence how people interpret risk. This study investigates how flood map types and sources influence individuals’ risk perception and decision-making. Using a randomized control trail (N = 796), participants were assigned to view one of the several flood risk representations sourced from governmental agencies, a nonprofit, and crowdsourced images of past floods, or plain geographical maps. Participants then self-reported their risk perceptions and behavioral intentions. This study compares visual formats and sources to reveal how communication strategies influence public understanding and preparedness, guiding more effective flood risk messaging.

This paper introduces ConStrobe – Construction Operations Simulation for Time and Resource Based Evaluations – which is a simulation software that builds upon knowledge in construction field operations simulation by providing the capabilities of running High Level Architecture (HLA)-compliant distributed simulations and being amenable to automation from external programs written in the Python language for two-way communication with external data sources. These features are provided to overcome some of the major limitations of existing construction operations simulation tools that have hindered their widespread adoption by industry. The framework of this software is explained along with a sample demonstration case to provide users with an overview of its capabilities and understanding of its working. It is anticipated that the novel capabilities of ConStrobe can reduce the time and effort required to create simulations to enable process analysis for decision-making under uncertainty for complex operations in the construction and built operations domain.

Precast concrete manufacturers increasingly face throughput bottlenecks as market demand rises and curing-area capacity reaches its limit. This paper develops a validated discrete-event simulation (DES) model of a Canadian precast panel plant using the Simphony platform. Field observations, time studies, and staff interviews supply task durations, resource data, and variability distributions. After verification and validation against production logs, two improvement scenarios are tested: (1) doubling curing beds and (2) halving curing time with steam curing. Scenario A reduces total cycle time by 26 %, while Scenario B achieves a 24 % reduction and lowers curing-bed utilization by 5 %. Both scenarios cut crane waiting and queue lengths, demonstrating that relieving the curing bottleneck drives system-wide gains. The study confirms DES as an effective, low-risk decision-support tool for off-site construction, offering plant managers clear, data-driven guidance for investment planning and lean implementation.

Computational Fluid Dynamics (CFD) is a widely used method for wind modeling in autonomous drone training, yet its computational expense limits real-time applications. This study explores the reliability of a convolutional autoencoder-based approach as a replacement for CFD in wind field generation. A convolutional autoencoder is trained on CFD-generated wind data to learn its spatial and dynamic distribution patterns, enabling the generation of realistic wind fields with significantly lower computational costs. The generated wind fields are then used to train reinforcement learning (RL)-based drones, with their policies evaluated in real CFD environments. Results demonstrate that the convolutional autoencoder accurately replicates CFD wind patterns, supports stable drone navigation, and facilitates seamless transferability of RL-trained policies from autoencoder-generated wind environments to CFD-based environments. These findings highlight the convolutional autoencoder’s potential as a computationally efficient and reliable alternative to traditional CFD in autonomous drone simulations.

As transportation systems move toward electrification and decarbonization, multifunctional charging stations (MFCS) are emerging as key infrastructure for electric vehicles (EVs) and hydrogen fuel cell vehicles (HFCVs). This paper presents a simulation model of an MFCS that integrates solar photovoltaic (PV), wind power, battery storage, hydrogen (H2) production, dual-pressure H2 storage, fuel cells, and dynamic grid interactions. The model simulates daily operations using 5-minute resolution data to capture realtime variability in renewable energy (RE), demand, and electricity prices. A flexible dispatch algorithm dynamically allocates energy for EV charging, H2 production, storage, and grid transactions while respecting system constraints. Results show that the MFCS effectively prioritizes RE usage, minimizes waste, meets diverse energy demands, and achieves net operational profit. The model serves as a valuable decision-support tool for designing and optimizing integrated clean energy hubs for zero-emission transportation.

Modern transportation research relies on seamlessly integrating traffic signal data with robust network representation and simulation tools. This study presents utdf2gmns, an open-source Python tool that automates conversion of the Universal Traffic Data Format, including network representation, signalized intersections, and turning volumes into the General Modeling Network Specification (GMNS) Standard. The resulting GMNS-compliant network can be converted for microsimulation in SUMO. By automatically extracting intersection control parameters and aligning them with GMNS conventions, utdf2gmns minimizes manual preprocessing and data loss. utdf2gmns also integrates with the Sigma-X engine to extract and visualize key traffic control metrics, such as phasing diagrams, turning volumes, volume-to-capacity ratios, and control delays. This streamlined workflow enables efficient scenario testing, accurate model building, and consistent data management. Validated through case studies, utdf2gmns reliably models complex urban corridors, promoting reproducibility and standardization. Documentation is available on GitHub and PyPI, supporting easy integration and community engagement.

Ensuring reliability of energy systems is critical for maintaining a secure and adequate energy supply, especially as the integration of renewable energy increases systems’ complexity and variability. Digital Twins offer a promising approach for data-driven reliability assessment and decision support in energy systems. Digital Twins provide decision support by dynamically modeling and analyzing system reliability using real-time data to create a digital replica of the physical counterpart. As modern energy systems generate vast amounts of data, it is essential to precisely define the data required for enabling Digital Twins for their reliability assessment. In this paper, we systematically investigate the data requirements for reliability-oriented Digital Twins for energy systems and propose a structured categorization of these requirements. To illustrate our findings, we present a case study demonstrating the link between data and model extraction for enhancing system reliability.

Fleet planning is often challenging. Especially, the dynamic interaction with maintenance entails various uncertainties. Predicting the arrivals into maintenance processes requires an understanding of fleet performance over time while, in turn, delays of repairs severely impact fleet performance and deterioration. This feedback loop has been neglected so far, which is why we present a novel framework using a ‘rolling window’ machine learning model to predict the inputs into a discrete event simulation (DES) of repair activities based on remaining useful life (RUL). Our ‘fleet tracker’ then uses the DES outputs to simulate fleet performance together with environmental and mission-based factors which form the inputs for predicting RUL. Finally, explainable ML helps decision-makers construct relevant ‘what if’ scenarios. As a motivating example, we consider helicopters in search and rescue missions and their maintenance. As a key result, we compare two scenarios of repair turnaround times and their impact on RUL decline.

Computationally expensive combat simulations are often used to inform military decision-making and sensitivity analyses enable the quantification of the effect of military capabilities or tactics on combat mission effectiveness. The sensitivity analysis is performed using a meta-model approximating the simulation's input-output relationship and the output data that most combat meta-models are fitted to correspond to end-of-run mission effectiveness measures. However during execution, a simulation records a large array of temporal data. This paper seeks to examine whether functional combat meta-models fitted to this temporal data, and the subsequent sensitivity analysis, could provide a richer characterization of the effect of military capabilities or tactics. An approach from Functional Data Analysis will be used to illustrate the potential benefits on a case study involving a closed-loop, stochastic land combat simulation.

The performance of a CPU-only implementation of the restarted GMRES algorithm with direct randomized-SVD -based preconditioning has been analyzed. The method has been tested on a set of sparse and dense matrices exhibiting varying spectral properties and compared to the ILU(0) -based preconditioning. This comparison aims to assess the advantages and drawbacks of both approaches. The trade-off between iteration-to-solution and time-to-solution metrics is discussed, demonstrating that the proposed method achieves an improved convergence rate in terms of iterations. Additionally, the method’s competitiveness with respect to both metrics is discussed within the context of several relevant scenarios, particularly those where GMRES-based simulation techniques are applicable.

Simulation software development is crucial for advancing digital engineering, scientific research, and innovation. The emergence of Generative AI, especially Large Language Models (LLMs), introduces possibilities for automating code generation, documentation, testing, and refactoring. Prompt engineering has become a key method of translating human intent into automated software output. However, integrating LLMs into software workflows requires reliable evaluation methods. This paper proposes a metric-based framework that uses software metrics to assess and guide LLM integration. Treating prompts as first-class artifacts, the framework supports improvements in code quality, maintainability, and developer efficiency. Performance, measured by execution time relative to a known high-performance codebase, is used as the initial metric to study. Work focuses on the impact of assigning roles in prompts and refining prompt engineering strategies to generate high-performance software through structured preambles. The work provides the foundation for LLM generated software starting from a well-defined simulation model.

Transitioning from narratives to formal system dynamics (SD) models is a complex task that involves identifying variables, their interconnections, feedback loops, and the dynamic behaviors they exhibit. This paper investigates how large language models (LLMs), specifically GPT-4o, can support this process by bridging narratives and formal SD structures. We compare zero-shot prompting with chain-of-thought (CoT) iterations using three case studies based on well-known system archetypes. We evaluate the LLM’s ability to identify the systemic structures, variables, causal links, polarities, and feedback loop patterns. We present both quantitative and qualitative assessments of the results. Our study demonstrates the potential of guided reasoning to improve the transition from narratives to system archetypes. We also discuss the challenges of automating SD modeling, particularly in scaling to more complex systems, and propose future directions for advancing toward automated modeling and simulation in SD assisted by AI.

Large language models (LLMs) have exhibited capabilities comparable to those of human experts in various fields. However, their modeling abilities—the process of converting real-life problems (or their verbal descriptions) into sensible mathematical models—have been underexplored. In this work, we take the first step to evaluate LLMs’ abilities to solve stochastic modeling problems, a model class at the core of Operations Research (OR) and decision-making more broadly. We manually procure a representative set of graduate-level homework and doctoral qualification-exam problems and test LLMs’ abilities to solve them. We further leverage SimOpt, an open-source library of simulation-optimization problems and solvers, to investigate LLMs’ abilities to make real-world decisions. Our results show that, though a nontrivial amount of work is still needed to reliably automate the stochastic modeling pipeline in reality, state-of-the-art LLMs demonstrate proficiency on par with human experts in both classroom and practical settings.

The advancement of Artificial Intelligence has accelerated the transformation of simulation from a tool for analysis into a dynamic partner in decision-making and system design. As AI systems become more capable of learning, reasoning, and adapting, simulation is evolving into an intelligent, autonomous, and predictive framework for exploring complex futures. This paper brings together future-oriented perspectives from simulation scientists and AI experts to discuss the next 25 years of innovation, collaboration, and disruption.

Modern simulation environments for complex multi-agent interactions must balance high-fidelity detail with computational efficiency. We present DECOY, a novel multi-agent simulator that abstracts strategic, long-horizon planning in 3D terrains into high-level discretized simulation while preserving low-level environmental fidelity. Using Counter-Strike: Global Offensive (CS:GO) as a testbed, our framework accurately simulates gameplay using only movement decisions as tactical positioning—without explicitly modeling low-level mechanics such as aiming and shooting. Central to our approach is a waypoint system that simplifies and discretizes continuous states and actions, paired with neural predictive and generative models trained on real CS:GO tournament data to reconstruct event outcomes. Extensive evaluations show that replays generated from human data in DECOY closely match those observed in the original game. Our publicly available simulation environment provides a valuable tool for advancing research in strategic multi-agent planning and behavior generation.

Addressing complex stochastic optimization problems often requires hybridization of search and evaluation methods. Simheuristics combine metaheuristics and simulation but typically rely on static control logic. Meanwhile, large language models (LLMs) offer advanced reasoning but lack robust mechanisms for constrained optimization. We propose the agentic simheuristic framework, a novel architecture that leverages an LLM as a high-level coordinator for simheuristic components. Applied to the team orienteering problem (TOP) under uncertainty, the framework employs an LLM to manage an exploratory agent for broad solution search and an exploitative agent for intensive refinement. Both agents integrate Monte Carlo simulation to evaluate solutions under uncertainty. The LLM guides the process by selecting promising and diverse exploratory solutions to seed refinement, enabling intelligent coordination within simheuristics. We present the framework architecture and provide initial empirical results on TOP benchmark instances, illustrating operational feasibility as a proof of concept and highlighting potential for explainable, AI-driven optimization.

This paper addresses the NP-hard cloud workflow scheduling problem by proposing a novel method that integrates Graph Neural Networks with Monte Carlo Tree Search (MCTS). Cloud workflows, represented as Directed Acyclic Graphs, present significant scheduling challenges due to complex task dependencies and heterogeneous resource requirements. Our method leverages Anisotropic Graph Neural Networks to extract structural features from workflow and create a heatmap that guides the MCTS process during both the selection and simulation phases. Extensive experiments on workflows ranging from 30 to 110 tasks demonstrate that our method outperforms rule-based algorithms, classic MCTS, and other learning-based approaches; more notably, it achieves near-optimal solutions with only a 2.56% gap from exact solutions and demonstrates exceptional scalability to completely unseen workflow sizes. This synergistic integration of neural network patterns with Monte Carlo simulation-based search not only advances cloud workflow scheduling but also offers valuable insights for simulation-based optimization across diverse domains.

In discrete event simulation (DES) of surgical centers, activity durations are inherently stochastic and are typically modeled with standard probability distributions. In reality, however, actual procedure times may not conform to standard distributions, depending instead on the individualized characteristics of patients. Accounting for such covariate effects in DES is challenging, making personalized care process simulation impractical. This shortcoming can compromise model fidelity and constrain the utility of DES for precise service management. Machine learning (ML) techniques offer a promising solution by incorporating patientspecific features into duration estimates. In this study, we introduce an ML-driven simulation framework that integrates advanced predictive algorithms with a DES model of a surgical center. We benchmark our approach against conventional DES methodologies that rely solely on predefined probability distributions to represent time variability. Results show that our approach can significantly reduce the randomness of service durations and lead to a significant improvement in simulation accuracy.

In this work, we introduce DES-Gymnax, a novel high-performance discrete-event simulator implemented in JAX. By leveraging the just-in-time compilation, automatic vectorization, and GPU acceleration capabilities of JAX, DES-Gymnax can achieve 10x to 100x times performance improvement over traditional PYTHONbased discrete-event simulators like Salabim. The proposed DES-Gymnax can feature a Gym-like API that facilitates seamless integration with reinforcement learning algorithms, addressing a critical gap between simulation engines and AI techniques. DES-Gymnax is validated on three benchmark models, i.e., an M/M/1 queue, a multi-server model, and a tandem queue model. Experimental results demonstrate that DES-Gymnax maintains simulation accuracy while significantly reducing execution time, enabling efficient large-scale sampling crucial for reinforcement learning applications in operations research areas. The open-source code is available in the DES-Gymnax repository (Yun, Jun, and Xiangfeng 2025).

We introduce a novel parallel discrete event simulation (PDES)-based method to couple multiple AI and non-AI agents in a rule-based manner with dynamic constrains while ensuring correctness of output. Our coupling mechanism enables the agents to work in a co-operative environment towards a common goal while many sub-tasks run in parallel. AI agents trained on vast amounts of human data naturally model complex human behaviors and emotions easily – this is in contrast to conventional agents which need to be burdensomely complex to capture aspects of human behavior. Distributing smaller AI agents on a large heterogeneous CPU/GPU cluster enables extremely scalable simulation tapping into the collective complexity of individual smaller models, while circumventing local memory bottlenecks for model parameters. We illustrate the potential of our approach with examples from traffic simulation and robot gathering, where we find our coupling of AI/non-AI agents improves overall fidelity.

Production planning and control (PPC) is challenged by the complex and volatile environment manufacturers face. One promising approach in PPC is the application of Reinforcement Learning (RL). In RL, an intelligent agent is trained in a simulation environment based on its experiences. The behavior of the agent is trained by defining a reward function that provides positive feedback if the agent performs well and negative feedback if it does not. Accordingly, the design of the reward function determines the impact RL can have. This article deals with the challenge of how to design a suitable reward function. To do so, 8 design principles and 21 design parameters were identified based on a structured literature review. The principles and parameters were utilized to systematically derive reward function alternatives for a given PPC task. These alternatives were applied to a use case in rough production scheduling being a sub task of PPC.

Industrial configuration planning requires testing many setups, which is time-consuming when each scenario must be evaluated through detailed simulation. To accelerate this process, we train a Multi-Layer Perceptron (MLP) to predict key performance indicators (KPIs) quickly, using it as a surrogate model. However, classical regression metrics such as Mean Squared Error (MSE), Mean Absolute Error (MAE), and Root Mean Squared Error (RMSE) do not reflect prediction quality in all situations. To solve this issue, we introduce a classification-based evaluation strategy. We define acceptable prediction margins based on business constraints, then convert the regression output into discrete classes. We assess model performance using precision and recall. This approach reveals where the model makes critical errors and helps decision-makers at Airbus Helicopters trust the AI’s predictions.

The Unrelated Parallel Machine Scheduling Problem (UPMSP) with release dates, setups, and eligibility constraints presents a significant multi-objective challenge. Traditional methods struggle to balance minimizing Total Weighted Tardiness (TWT) and Total Setup Time (TST). This paper proposes a Deep Reinforcement Learning framework using Proximal Policy Optimization (PPO) and a Graph Neural Network (GNN). The GNN effectively represents the complex state of jobs, machines, and setups, allowing the PPO agent to learn a direct scheduling policy. Guided by a multi-objective reward function, the agent simultaneously minimizes TWT and TST. Experimental results on benchmark instances demonstrate that our PPO-GNN agent significantly outperforms a standard dispatching rule and a metaheuristic, achieving a superior trade-off between both objectives. This provides a robust and scalable solution for complex manufacturing scheduling.

This paper investigates the sample efficient exploration policy for asynchronous Q-learning from the perspective of uncertainty quantification. Although algorithms like $\epsilon$-greedy can balance exploration and exploitation, their performances heavily depend on hyperparameter selection, and a systematic approach to designing exploration policies remains an open question. Inspired by contextual Ranking and Selection problems, we focus on optimizing the probability of correctly selecting optimal actions (PCS) rather than merely estimating Q-values accurately. We establish a novel central limit theorem for asynchronous Q-iterations, enabling the development of two strategies: (1) an optimization-based policy that seeks an optimal computing budget allocation and (2) a parameter-based policy that selects from a parametrized family of policies. Specifically, we propose minimizing an asymptotic proxy of Q-value uncertainty with regularization. Experimental results on benchmark problems, including River Swim and Machine Replacement, demonstrate that the proposed policies can effectively identify sample-efficient exploration strategies.

As artificial intelligence rapidly co-evolves with complex modern systems, new simulation frameworks are needed to explore the potential impacts. In this article, I introduce a novel open source multi-agent financial market simulation powered by raw historical order streams at nanosecond resolution. The simulation is particularly targeted at deep reinforcement learning, but also includes momentum, noise, order book imbalance, and value traders, any number and type of which may simultaneously trade against one another and the historical order stream within the limit order books of the simulated exchange. The simulation includes variable message latency, automatic agent computation delays sampled in real time, and built-in tools for performance logging, statistical analysis, and plotting. I present the simulation features and design, demonstrate the framework on a multipart DeepRL use case with continuous actions and observations, and discuss potential future work.

Lifelong deep learning aims to enable neural networks to continuously learn across tasks while retaining previously acquired knowledge. This paper introduces an algorithm, Task-Aware Multi-Expert (TAME), which facilitates incremental and collaborative learning by leveraging task similarity to guide expert model selection and knowledge transfer. TAME retains a pool of pretrained neural networks and selectively activates the most relevant expert based on task similarity metrics. A shared dense layer then utilizes the appropriate expert's knowledge to generate a prediction. To mitigate catastrophic forgetting, TAME incorporates a limited-capacity replay buffer that stores representative samples and their embeddings from each task. Furthermore, an attention mechanism is integrated to dynamically prioritize the most relevant stored knowledge for each new task. The proposed algorithm is both flexible and adaptable across diverse learning scenarios. Experimental results on classification tasks derived from CIFAR-100 demonstrate that TAME significantly enhances classification performance while preserving knowledge across evolving task sequences.

Simulating realistic populations for strategic influence and social-cyber modeling requires agents that are demographically grounded, emotionally expressive, and contextually coherent. Existing agent-based models often fail to capture the psychological and ideological diversity found in real-world populations. This paper introduces AURORA, a Retrieval-Augmented Generation (RAG)-enhanced framework that leverages large language models (LLMs), semantic vector search, and salience-aware topic modeling to construct synthetic communities and personas. We compare two opinion modeling strategies and evaluate three LLMs—gemini-2.0-flash, deepseek-chat, and gpt-4o-mini—in generating emotionally and ideologically varied agents. Results show that community-guided strategies improve meso-level opinion realism, and LLM selection significantly affects persona traits and emotions. These findings demonstrate that principled LLM integration and salience-aware modeling can enhance the realism and strategic utility of synthetic populations for simulating narrative diffusion, belief change, and social response in complex information environments.

Generative-AI models offer powerful capabilities for learning complex dynamics and generating high-fidelity synthetic data. In this work, we propose Conditional Temporal Diffusion (CTD) models for generating wafer fabrication time-series trajectories conditioned on static factory configurations. The model is trained using data from a Parallel Discrete Event System Specification (PDEVS)-based MiniFab benchmark model, which simulates different steps of a semiconductor manufacturing process and captures the wafer processing dynamics (e.g., throughput \& turnaround time). These simulations incorporate multiscale, realistic behaviors such as preventive maintenance and wafer dispatching under both uniform and sinusoidal generation patterns. CTD models are conditioned on static covariates, including wafer composition, lot sizes, repair type, and wafer generator mode of the factory. Experimental evaluations demonstrate that the synthetic outputs achieve high fidelity with average errors below 15\% while significantly reducing data generation time. This highlights CTD’s effectiveness as a scalable and efficient surrogate for complex manufacturing simulations.

Input modeling is pivotal for generating realistic data that mirrors real-world variables for simulation, yet traditional parametric methods often fail to capture complex dependencies. This study investigates the efficacy of modern generative models—such as Generative Adversarial Networks (GANs), Variational Autoencoders (VAEs), Normalizing Flows, and Diffusion Models—in addressing these challenges. Through systematic experiments on synthetic and real-world datasets, we evaluate their performance using metrics like Wasserstein distance and quantile loss. Our findings reveal that VAEs and Denoising Diffusion Probabilistic Models (DDPMs) consistently outperform other models, particularly in capturing nonlinear relationships, while GAN-based approaches exhibit instability. These results provide practical insights for practitioners, highlighting models that deliver reliable performance without extensive customization, and outline promising directions for future research in simulation input modeling.

Genetic Programming (GP) is a powerful evolutionary computation technique for automatic program synthesis, but it suffers from inefficient search and solution bloat. This paper proposes using Monte Carlo Tree Search (MCTS) to directly construct programs by formulating synthesis as a sequential decision problem modeled as a Markov Decision Process. We tailor the MCTS algorithm to the GP domain and demonstrate its theoretical consistency. Empirical evaluation on symbolic regression benchmarks shows that our approach consistently outperforms traditional GP, discovering higher-quality solutions with more compact structural complexity and suggesting that MCTS is a promising alternative for automatic program generation.

Modern industrial control systems (ICSs) are increasingly relying upon IoT and CPS technology to improve cost-effective service performance at scale. Consequently, the cyber vulnerability terrain is largely amplified in ICSs. Unfortunately, the historical lack of (a) sufficient, non-noisy ICS cyber incident data, and (b) intelligent operational business processes to collect and analyze available ICS cyber incident data, demands the attention of the Bayesian AI community to develop cyber risk management (CRM) tools to address these challenges. In this paper we show with sufficient Monte Carlo simulation evidence that Bayesian AI on noisy (and small) ICS cyber incident data is ineffective for CRM. More specifically, we show via a novel graphical sensitivity analysis methodology that even small amounts of statistical noise in cyber incident data are sufficient to reduce ICS intrusion/anomaly detection performance by a significant percentage. Hence, ICS management processes should strive to collect sufficient non-noisy cyber incident data.

Cyber insurance (CI) markets improve the cybersecurity of companies (enterprises) in theory. This theory is increasingly being validated in practice in multiple enterprises that buy cyber insurance. However, the supply-demand gap in the CI market is huge and stretches over a trillion dollars. The primary reason being that cyber reinsurance companies do not have sufficient capital to serve their CI company clients through profitable cyber-risk portfolio diversification. This capital problem is subsequently rooted in the fundamental challenge of enterprise cyber posture information asymmetry that has plagued the CI industry since its inception. A radical capital-boosting mechanism proposed by researchers and deployed within the industry in the last two years is an insurance-linked security (ILS) product such as a catastrophic (CAT) bond. In this paper, we present arguments on why, how, and when CAT bonds help sustain capital-boosted reinsurance markets complemented by innovative AI/ML-driven inside-out cyber posture scanning solutions.

The Information Universe (IU) communicates with the Material Universe (MU) to create and repair atoms. This is required because quarks and bosons that make up atoms have a relatively short life and must be replaced. The communication messages are described by a context-sensitive language specified using message generating rules. The SUSY (Supersymmetric) inversion model is a process defined by these rules that describes how subatomic particles are made and combined to create or repair atoms; indeed, there is a language message (a sequence of process actions) for every IU/MU system regulatory problem. An OpEMCSS (Operational Evaluation Model for Complex Sensitive Systems) simulation model of the IU/MU system can learn these rules to gain an understanding of the SUSY messaging process. The IU/MU system simulation model will be used to learn and generate messages that result in the LENR (Low Energy Nuclear Reaction) system producing useful new physics.

This paper presents a simulation-driven development approach for a camera-based anti-collision system designed for automated tower cranes. Stereo camera systems mounted directly on the crane's hook generate real-time 3D point clouds to detect people in the immediate danger zone of suspended loads. A virtual construction site was implemented in a game engine to simulate dynamic scenarios and varying weather conditions. The system utilizes a neural network for pedestrian detection and computes the minimum distance between load and detected persons. A closed-loop architecture enables real-time data exchange between simulation and processing components and allows easy transition to real-world cranes. The system was evaluated under different visibility conditions, showing high detection accuracy in clear weather and degraded performance in fog and rain due to the limitations of stereo vision. The results demonstrate the feasibility of using synthetic environments and point cloud-based perception to develop safety-critical assistance systems in construction automation.

Dynamic job shop scheduling (DJSS) demands rapid, data-driven decisions to balance conflicting objectives in complex, stochastic manufacturing environments. While Artificial Intelligence, particularly deep reinforcement learning (RL), offers powerful optimization capabilities, its development and evaluation are often limited by the lack of suitable high-speed, flexible simulation testbeds. To address this, we introduce FabSim, a micro-discrete-event simulation (micro-DES) environment purpose-built for developing and evaluating intelligent scheduling strategies for DJSS. FabSim models core scheduling dynamics, including resource contention, flexible routing, batching, and stochastic events, using an efficient event-driven kernel and indexed look-up tables that enable sub-second simulation runtimes. Compliant with the Farama Gymnasium API, it integrates seamlessly with standard machine learning libraries. FabSim facilitates reproducible benchmarking and serves as a practical back-end for digital twin applications requiring near-real-time analysis. This paper details FabSim's design, validates its high-speed performance, and demonstrates its utility for AI-driven scheduling research.

This paper addresses the challenge of detecting heavy congestion in pedestrian crowds. While congestion detection is straightforward when monitoring predefined locations through pedestrian density measurements, identifying potential congestion areas in advance remains problematic. We propose a novel algorithm that eliminates the need for prior knowledge of congestion-prone locations by topological data analysis. Our approach transforms time-series pedestrian position data into topological representations, enabling more robust congestion detection. We validate our algorithm using both synthetic data from multi-agent simulations and real-world pedestrian measurements. The experimental results demonstrate that converting traditional positional data into topological representations significantly improves the performance of machine learning models in congestion detection tasks.

How can a traffic simulation be designed to faithfully reflect real-world traffic conditions? One crucial step is modeling the volume of traffic demand. But past demand modeling approaches have relied on unrealistic or suboptimal heuristics, and they have failed to adequately account for the effects of noisy and multimodal data on simulation outcomes. In this work, we integrate advances in AI to construct a three-step, end-to-end pipeline for systematically modeling traffic demand from detector data: computer vision for vehicle counting from noisy camera footage, combinatorial optimization for vehicle route generation from multimodal data, and large language models for iterative simulation refinement from natural language feedback. Using a road network from Strongsville, Ohio as a testbed, we show that our pipeline accurately captures the city’s traffic patterns in a granular simulation. Beyond Strongsville, incorporating noise and multimodality makes our framework generalizable to municipalities with different levels of data and infrastructure availability.

Simulations are widely used to evaluate public health interventions, yet they often fail to quantify how interventions in one region indirectly affect others—a phenomenon known as spillover. This omission can lead to incorrect policy evaluations and misattributed effects. We propose a post-simulation framework for estimating causal spillover in spatial epidemic networks. Our method introduces a directional graph neural network (Dir-GNN) estimator that learns homophily-aware representations and estimates counterfactual outcomes under hypothetical neighbor treatments. Applied to a semi-synthetic setup built on PatchSim—a metapopulation SEIR simulator with realistic inter-county mobility—our estimator recovers spillover effects and corrects attribution errors inherent in standard evaluation. Experiments show that accounting for spillover improves treatment estimation and policy reliability.

Discrete-event simulation is widely used in the healthcare field to optimize processes and manage resources. This study presents a DES model developed for a molecular biology laboratory in Colombia, recognized by Oxford Nanopore Technologies and specializing in Sanger and Nanopore sequencing techniques. Using real data, the model analyzes processing times for each sequencing method and estimates the laboratory’s maximum capacity under varying technician experience, equipment availability, task durations and number of samples processed. The goal is to provide genomic laboratories and regulatory stakeholders with a flexible tool to evaluate performance, the impact of alternative configurations, and support capacity planning under resource and demand constraints.

This study presents the applications, challenges, and future research directions of simulation for health inequalities research based on a systematic literature review and survey of authors included in the review.

Burnout among healthcare workers in inpatient care is worsened by administrative tasks and staff shortages. Although service robots could support these environments, adoption remains limited, especially for non-clinical roles, and they lack structured methods for early-stage evaluation. This paper presents a conceptual hybrid simulation model as a foundation for pre-adoption analysis of robotic agents in inpatient care. The model integrates Discrete Event Simulation for patient flow and Agent-Based Modeling for human–robot interactions, designed using the ODD protocol and informed by expert input. It formalizes assumptions, roles, and workflow logic to support early experimentation and scenario testing. Initial simulations assess task distributions with and without robot integration, highlighting reduced documentation and assistance burdens for nurses. The conceptual model was validated through structured meetings with clinical staff and proved feasible for further development. The model includes the assumptions, structure, and logic behind the simulation, serving as a basis for future model expansion.

Over the past few years, the interest in Machine Learning (ML) has grown due to its ability to improve solutions related to other fields. This paper explores the use of ML techniques in simulation through a systematic literature review of the Winter Simulation Conference proceedings from 2013 to 2023. Our research is focused on the Discrete-Event Simulation (DES) field, centering our attention on the Discrete-Event System Specification (DEVS) formalism as a particular case. The research questions were designed to examine the most frequent contexts, applications, methods, and software tools used in these studies. As a result, this review reports insights into 44 research studies. The main contribution of this paper is related to systematically gathering, analyzing, and discussing the knowledge disseminated in these two areas (ML and DES), aiming to support future research and expand the literature in this field.

Road traffic management enters a new era with the automatic collection and analysis of big data. The traffic data could be collected continuously using various IoT sensors (light, video, etc.) and stored in the cloud. The collected data are then analyzed within a short time to make traffic control decisions, e.g., traffic redirection, traffic light duration change, and vehicle route recommendation. This study proposes (1) a traffic simulation considering a road network with several traffic lights and (2) regression machine learning models to understand the behavior of the vehicles based on the real-time characteristics of the traffic. The numerical experiment results show that (1) the best models are OrthogonalMatchingPursuitCV and the HuberRegressor, and (2) the road network behavior is affected by the condition of all intersections rather than only certain intersections or surrounding road segments.

The design of distributed control strategies for multi-robot systems (MRS) relies heavily on simulations to validate algorithms prior to real-world deployment. However, simulating such systems poses significant challenges due to their dynamic network topologies and scalability requirements, where full inter-robot communication becomes computationally prohibitive. In this paper, we extend the applications of the Emergent Behavior DEVS (EB-DEVS) formalism by developing an agent-based model (ABM) to address key distributed control challenges in networked MRS. The proposed approach supports both direct and indirect interactions between agents (robots) via event messages and through macroscopic-microscopic states sharing, respectively. We validate the model using a challenging cooperative target-capturing scenario that demands dynamic multi-hop communication and robust coordination among agents. This complex use case highlights the strengths of EB-DEVS in managing asynchronous events while minimizing communication overhead. The results demonstrate the formalism’s effectiveness in supporting decentralized control and simulation scalability within a hierarchical micro-macro modeling framework.

Belbin’s team role theory identifies nine behavioral roles that, when combined, support effective collaboration. Configuring teams based on these roles is often manual, costly, and inflexible. This article presents an individual-oriented simulation model using the Discrete-Event System Specification (DEVS) formalism to emulate group interactions shaped by Belbin roles. Each team member is modeled as an atomic entity with behavior defined by a combination of two roles. This enables controlled experimentation with different team compositions, interaction timings, and communication sequences. Simulations were conducted using synthetic data, defined under plausible assumptions based on Belbin’s framework. The model enables exploration of how different configurations affect communication flow and task distribution, supporting the identification of team structures that promote balance and efficiency. Results demonstrate the potential of integrating behavioral theories with formal modeling approaches to improve team design. This work offers a flexible and extensible simulation-based method for analyzing and optimizing team dynamics.

Routed Discrete Event System Specification (RDEVS) is a formalism that extends the Discrete Event System Specification (DEVS) to support the Modeling and Simulation (M&S) of routing processes in discrete event-based scenarios. We present a web-based tool that, following the principles of MDE, combines textual modeling with code generation to reduce development time, ensure consistency with the RDEVS formalism, and promote independence from external software tools.

This work presents a methodology for developing embedded applications in Internet-of-Things (IoT) and robotic systems through Model and Simulation (M\&S)-based design. We introduce adaptations to the PowerDEVS toolkit's abstract simulator to enable embedded execution on resource-constrained platforms, specifically targeting the widely used ESP32 development kit tailored to IoT systems. We present a library of DEVS atomic models designed for simulation-environment interaction, enabling embedded software development through sensor data acquisition and actuator control. To demonstrate the practical utility of the embedded PowerDEVS framework, we evaluate its performance in real-world discrete-event control applications, including a line-follower robot and an electric kettle temperature regulator. These case studies highlight the approach’s versatility and seamless integration in IoT and robotic systems.

The decarbonisation of industrial clusters is central to the UK’s net zero strategy. Located in southeast England, the Kemsley Industrial Cluster (comprising five firms) is the largest industrial cluster in the counties of Kent and Sussex, offering a critical opportunity for coordinated action. This study supports their decarbonisation-focused decision-making through a Monte Carlo simulation model. The model takes hourly energy use as input and simulates the potential impact of emerging decarbonisation technologies including hydrogen, and collaborative energy exchanges such as private wire electricity and shared steam flows. Key model outputs include emissions profiles and the overall energy balance. The model produces insights for six main decarbonisation scenarios over a 25-year horizon (2025–2050). Results suggest the cluster could reach net zero emissions by 2035 if carbon capture is implemented. This study develops a novel Monte Carlo model that captures inter-firm dynamics within industrial clusters, directly informing collaborative decarbonisation investment strategies.

This study presents a case of a packaging ink manufacturer based in Surabaya, Indonesia. The production process involves resources, including bowls, field operators, lab technicians, mixers, quality checking equipment, and a high-stacker. The company faces a persistent problem of low on-time delivery performance, which risks customer loss. To address this issue, a discrete-event simulation approach is employed, considering the system’s complexity, process interdependencies, and queuing dynamics. The simulation evaluates several improvement scenarios to identify the best alternative for enhancing on-time delivery while accounting for changes in variable costs. Additionally, analysis for Net Present Value (NPV) is used to assess the financial feasibility of selected solution alternatives. The results indicate that a rearrangement of production resources can significantly improve on-time delivery performance. Although variable costs increase, the selected alternative yields a positive NPV, justifying the investment. This study demonstrates how simulation-based decision-making can support resource planning in complex manufacturing environments

Airports function as complex systems comprising interrelated entities, resources, and processes. Inefficiencies within these systems, particularly long waiting times, contribute to passenger dissatisfaction. This study examines operational improvements at Mumbai International Airport, one of India’s busiest hubs, focusing on identifying bottlenecks and reducing congestion for departing passengers. Discrete event simulation served as the core methodological framework. Through hybrid simulation, AS-IS models were developed to analyze operational processes and evaluate bottlenecks. BPMN was used for conceptual modeling, followed by Simul8 for dynamic modeling. TO-BE system configurations for check-in, security screening and boarding were then remodeled. Simulation experiments were conducted to determine the optimal setup that minimizes queuing time while maximizing passenger throughput. The findings highlight an ideal system configuration that significantly reduces waiting times and overall passenger processing time, resulting in improved operational efficiency. These insights provide data-driven recommendations for optimizing airport processes, ultimately enhancing passenger experience and improving airport performance.

High volumes, short buffer times and many players influence the maritime transport chain and intra-port container transport. Congestion at logistics hubs leads to inefficient processes, increased waiting times, costs and emissions. Truck appointment systems are intended to reduce congestion by offering trucking companies time slots for delivering or collecting containers. However, these systems lack flexibility, which limits utilization and reduces efficiency. The study aims to develop strategies to evaluate and improve these systems in a simulation study. The results show that medium-length time slots improve the situation for transparent processes, or longer ones for limited transparency. Overbooking and open access are mutually exclusive, but offer slight advantages individually. Adjustments to opening hours of other nodes are show little benefit. Overall, flexibility options positively influence the proportion of load trips and waiting times.

This paper presents a general framework for planning and dimensioning a multi-service Internet Protocol (IP) network supporting voice, video, and data. We describe a methodology for dimensioning link capacity and packet delay while meeting Quality of Service (QoS) requirements under general traffic distributions, using discrete event simulation. Additionally, the paper proposes a procedure to optimize transmission probabilities between nodes, aiming to increase throughput and reduce delay based on the offered network traffic. This procedure applies different transmission probabilities for each node type. Admission control is implicitly incorporated to regulate the number of active streams. A Jackson network is employed to validate the simulation model. Once validated, other probability distributions are explored to assess achievable delay and throughput, and the resulting suboptimal optimization outcomes are presented.

This study aimed to conduct discrete-event simulations using Simul8 in key locations within the city of São Caetano do Sul, Brazil (shopping mall; supermarkets; gas stations) to assess the current level of service provided by electric vehicle charging stations. Simulations are also carried out to evaluate future scenarios, considering the projected growth of electric vehicles both nationally and within the city. In partnership with municipal departments, this work is providing valuable insights to the city's master plan, currently under development, which will outline strategic guidelines for urban growth over the next ten years. In addition to assessing current infrastructure performance, the study seeks to determine the optimal number of charging stations required to ensure a satisfactory level of service under different future demand scenarios.

We consider the problem of Automated Model Generation (AMG) for Digital Twins (DTs) of manufacturing systems, particularly those represented as stochastic Discrete-Event Simulation (DES) models. Unlike fully data-driven approaches, where both model parameters and structure are inferred from event logs, we propose an expert-in-the-loop approach targeting systems whose structure changes only occasionally, with such changes being automatically detected. While machine states and parameters (such as task delays) are continuously inferred from data, the process remains expert-tunable via a GUI-based, guided flow. A natural language description of system structure is translated into readable DES models using LLMs. The model is built as an interconnection of configurable components from a lightweight, open-source library FactorySimPy, with parameters inferred seamlessly from data. We outline the proposed flow, its components, and results from a proof-of-concept implementation, and provide a detailed review of existing AMG approaches, highlighting key differentiating aspects of our framework.

To remain competitive in an evolving market, enterprises must adopt modern approaches in their production lines. Flexible Manufacturing Systems (FMS) produce diverse, high-quality products with short processing times. New technologies have transformed manufacturing. Among them, Digital Twin (DT) technology improves decision-making through real-time simulations. Most studies on FMS focus on reducing makespan. This paper proposes a model optimizing the makespan and energy consumption and production costs. It also presents a DT framework with a physical and virtual part connected in real time. The framework includes production data, optimization, scheduling, and learning, etc. Two experiments are conducted. The first uses Simulated Annealing (SA) to minimize makespan. Results show that SA is flexible, finding different schedules with the same makespan. The second applies Archived Multi-Objective Simulated Annealing (AMOSA) to optimize makespan, production cost, and energy consumption. Results show that AMOSA provides better trade-offs between objectives, making it effective for complex FMS.

This work explores the enhancement of the Earmalean line in order to increase accessibility, operational efficiency, and educational impact. FlexSim is used to model system behavior, process interactions, and human-machine collaboration in a dynamic and visual environment. The simulation incorporated realistic process logic, including part flows, workstation interactions, stock management, and operator tasks, allowing the testing of multiple what-if scenarios under controlled conditions. FlexSim’s 3D modeling capabilities and integrated analytics tools enabled the evaluation of key performance indicators such as cycle time, throughput, resource utilization, and work-in-process inventory. Data-driven modeling enabled realistic scenario testing, including automated logistics and human-robot collaboration. The results underline the value of integrating smart technologies into didactic systems and open pathways for future development toward a connected and adaptive learning platform.

This study addresses the problem of optimal sensor placement in discrete-event systems modeled by partially observable Petri nets, with the dual objectives of ensuring causal observability and minimizing total installation cost. The system is algebraically reformulated as a descriptor system, which leads to a combinatorial, nonlinear, and non-convex optimization problem. To overcome these difficulties, two metaheuristic approaches, simulated annealing and genetic algorithms, are proposed and evaluated on a complex manufacturing system. The results obtained highlight the effectiveness and scalability of both methods, underscoring their strong potential for industrial applications.

Layout decisions in empty-container depot affect both efficiency and safety, but prior studies rarely evaluate them together. This work presents an integrated discrete-event simulation framework that quantifies truck turnaround times, lifting equipment travel distances, and potential collision risks. A case study in a Chilean multipurpose terminal shows the value of the proposed framework, revealing critical workload imbalances and traffic bottlenecks, particularly for reefer operations. Statistical analyses show significant performance degradation under high truck arrival frequencies and confirm that traditional efficiency metrics only partially explain safety risks. Managerial insights include layout redesign, targeted equipment allocation, and automated gate operations. The study highlights the necessity of balancing operational efficiency and safety in space-constrained terminals to improve overall performance.

Inventory management needs coordinated efforts among internal and external stakeholders. An electric power distribution service unit faces challenges in the inventory management. Operational constraints include ordering cycle, fluctuating lead times, and the absence of buffer stock must be managed while adhering to mandatory service standards. This study adopts a system dynamics simulation to identify key stakeholders and map their interdependencies. The objects of study are power cable, cubicle, insulator, and lightening arrester. A causal loop diagram (CLD) is employed to visualize and analyze the dynamics of the problem. Inventory system is modeled into ten sub models. It is material request sub model, purchasing order, order receipt, material issued transaction, inventory level, coordination, service level, inventory turnover, budget - total cost, and service duration day. Continuous review policy is integrated in the model as proposed policy and combination with supplier improvements gives best results in 26,12% increase in service level.

Post-consumer packaging waste continues to rise, intensifying the need for automated sorting to increase recycling efficiency. Lightweight packaging (LWP), with its variable geometries, materials, and occlusions, remains especially difficult for conventional vision systems. Integrating You Only Look Once (YOLO) instance segmentation into robotic simulation platforms enables robust real-time detection. Experiments show that synthetic datasets yield consistently high segmentation accuracy, whereas real-world performance fluctuates. Crucially, increasing model size or resolution does not guarantee improvement; task-specific tuning and system-level integration are more effective. Simulation frameworks combining Unity, Robot Operating System 2 (ROS2), and MoveIt2 provide realistic evaluation and optimization. These findings demonstrate that AI-based segmentation and digital twins can deliver scalable, adaptive, and self-optimizing sorting systems, offering a practical pathway to sustainable material recovery and circular economy implementation.

Leveraging Digital Twins, as near real-time replicas of physical systems, can help identify inefficiencies and optimize production in manufacturing systems. Digital Twins’ effectiveness, however, relies on continuous validation of the underlying models to ensure accuracy and reliability, which is particularly challenging for complex, multi-component systems where different components evolve at varying rates. Modular validation mitigates this challenge by decomposing models into smaller sub-models, allowing for tailored validation strategies. A key difficulty in this approach is preserving the interactions and dependencies among the sub-models while validating them individually; isolated validation may yield individually valid sub-models while failing to ensure overall model consistency. To address this, we build on our previously proposed modular validation framework and introduce an approach that enables sub-model validation while maintaining interdependencies. By ensuring that the validation process reflects these dependencies, our method enhances the effectiveness of Digital Twins in dynamic manufacturing environments.

This study presents an approach to dynamically calibrate a digital twin to support decision-making in systems operating under uncertainty. The framework integrates uncertainty by turning a physics-based model into a stochastic simulation, where independent variables that represent environmental conditions may be nonstationary whereas target variables are conditionally stationary. Calibration itself is formulated as a simulation optimization problem that we solve using a root-finding strategy. As a case study, we apply the framework to the prediction of short-term power deficit, known as the wake effect, in wind farms using real-world data and demonstrate the robustness of the proposed framework. Besides advancing the digital twin research, the presented methodology is expected to impact wind farm wake steering strategy by enabling accurate short-term wake effect prediction.

This paper introduces fireSCore, an open source framework for incident risk evaluation, simulation, coverage optimization, and relocation experiments. As a digital twin of operational fire department logistics, its visualization frontend provides a live view of current coverage for the most common fire department units. Manually changing a unit status allows for a view into future coverage as it triggers an immediate recalculation of prognosed response times and coverage using the Open Source Routing Machine. The backend provides the controller and model, and implements various algorithms, e.g. a relocation algorithm that optimizes coverage during major incidents. The data broker handles communication with data sources and provides data for the front- and backend. An optional simulator adds an environment in which various scenarios, models and algorithms can be tested and aims to drive current and future organizational developments within the Dutch national fire service.

Urban Air Mobility presents unique community impacts. UAM vehicles mostly operate above populated areas while traditional aviation typically operates point-to-point between populated areas, so impacts on third parties are increasingly important. This paper explores the approach of using a Digital Twin to optimize EVTOL vehicle flight planning by using accurate live population data to minimize third-party safety and privacy impacts. Live population density data is translated into an equivalent agent-based simulation and used to calculate safety and privacy metrics. A Montreal vertiport case study compares a baseline EVTOL approach suggested by the FAA to 128 alternate approach scenarios to find generalizations and prove the usefulness of Digital Twin technology for UAM operational optimization. It was found that flight path characteristics suggested by regulators are not necessarily optimal when considering third-party impacts, and by extension that Digital Twins are promising technology that will play a significant role in making UAM safer.

Digital twins (DT) are increasingly being adopted to improve system monitoring, prediction, and decision making in various domains. Although simulation plays a central role in many DT implementations, a lack of formal modeling foundations often leads to ad hoc and non-scalable solutions. This paper proposes a simulation engine for DT applications based on the Discrete Event System Specification (DEVS) formalism. DEVS provides a robust, modular, and hierarchical modeling framework suitable for modeling the structure and behavior of complex cyber-physical systems. A key contribution is the integration of the Parallel DEVS for Multicomponent Systems (multiPDEVS) formalism with X-Machines to support state and memory separation for simulation models with the goal of improving model scalability and reusability, as well as providing a basis for integration with DTs. The paper presents the architectural design of the engine, highlights its main functional components, and demonstrates its capabilities using a preliminary use case.

This work develops a simulation engine to create a digital twin that will monitor a gambler’s betting behavior when playing games. The digital twin is designed to perform simulations to compare outcomes of different betting strategies, under various assumptions on the psychological profile of the player. With these simulations it then produces recommendations to the player aimed at mitigating adverse outcomes. Our work focuses on efficient simulation and the creation of the corresponding GUI that will become the interface between the player and the digital twin.