WSC 2004

WSC 2004 Final Abstracts

Simulation Case Studies Track

Sunday 1:00:00 PM 2:30:00 PM
Cases I

Chair: Brett Peters (Texas A&M University)

Route Planning and Threat Avoidance through Cognitive Robotics
Brad Best (Micro Analysis & Design)

This simulation case study demonstrates the use of the ACT-R cognitive modeling framework for robotic control in a real environment and virtual avatar control in a virtual environment. The model performs spatial tasks ranging from simple mobility and obstacle avoidance to more complex behaviors such as route planning while avoiding threats.

Utilizing Witness Simulation To Model The UH60 Blackhawk Production
Christopher S. Legge (Lanner Group Inc.)

A case study presentation on the development of a WITNESS / SIMBA simulation to assist the US ARMY monitor the UH60M Blackhawk manufacturing program, the development of a dashboard and the use of a Visual Basic application with embedded WITNESS Simulation Engine.

Sunday 3:00:00 PM 4:30:00 PM
Cases II

Chair: Jeffrey Smith (Auburn University)

Simulation Shows Savings with RFID Implementation
Carley Jurishica and Robert Schwieters (Rockwell Automation)

This presentation will focus on how simulation was used in supporting the business case for RFID. Recently, a pilot study at several Allen-Bradley facilities evaluated RFID (radio frequency identification) use in the production and distribution of circuit boards. The pilot consisted of RFID tag application to three different products manufactured in the Allen-Bradley Twinsburg Ohio facility. The attached RFID tags were read upon arrival to the Champaign Distribution Center (CDC). While the tag read percentages and other system variables could be explicitly measured, the impact of a large-scale RFID implementation remained unclear. In order to provide answers to the “what if…” questions proposed by the pilot project team, a simulation study of the RFID facility was launched. The study assisted in defining the impact on the labor savings and helped analyze the potential return on investment (ROI).

Analysis of Full Pallet Conveyor System
Darrell W Starks (Rockwell Automation) and John Fortini and Robert Schlafer (PepsiAmericas)

PepsiAmericas was experiencing throughput problems with its full pallet conveyor system at one of its bottling facilities. This system takes the full pallets of cans or bottles from the palletizers and sends them to the stretchwrappers and then into the warehouse. One solution was thought to be the addition of a stretchwrapper to the system. Before purchasing this equipment, PepsiAmericas decided to perform a quantitative analysis of the system using a simulation model. This model showed that a third stretchwrapper was not required. By changing a portion of the logic governing the movement of pallets from the five palletizers to the two current stretchwrappers, it was shown that the current two stretchwrappers could handle the throughput of the palletizers.

Monday 10:30:00 AM 12:00:00 PM
Cases III

Chair: Osman Balci (Virginia Tech)

This case study focuses on simulating a 65-node private Internet Protocol (IP) network and calibrating it with an analytical model of the same network. It provides an overview of the OPNET Modeler simulation tool and a discussion of the modeling process, as well as model results and analysis. Additionally, it presents comparisons of the results between the simulation model and the analytical model. Finally, it shares some important lessons learned about the modeling process and the OPNET modeling tool.

Combining Optimization and Simulation in Supply Chain Models
Andy Siprelle and Mike Denzien (Simulation Dynamics, Inc)

This case study presents how we have combined simulation with optimization to address problems in supply chain design. This case study illustrates the feasibility and benefits of combining optimization with simulaiton. We also present applications which employ dynamic programming and other mathematical programming techniques to determine where to place inventory in the supply chain.

VIP Protection Simulation Of Imaging In Concrete Realizing The Lagrangian Deformable Tensor Software(LDTS) Protocol For Homeland Defense And Security
Carol Ann Niznik (NW SYSTEMS)

Each VIP Protection simulation case study of imaging in concrete requires the clock timed catastrophe theory gateway networking software linkage between the Global Information Grid(GIG), i.e., the CIA Simulation facility, and the Lagrangian Deformable Tensor Software(LDTS) Protocol hardware sensors, detecting the motion in explosive containers within concrete. The queer differential equation in the LDTS Protocol, an astrophysics theoretical tool, can be optimized with respect to delay, weight, motion and perturbation sensitivity. The object oriented code for the LDTS Protocol is comprised of three sections for the Lagrangian Deformable Tensor within the top section of the Geometric Asscher Cut Structure(GACS). The four isosceles triangles on top of the GACS, represent the perturbation sensitivity, game theory, image processing and delay optimization parameters. The equilateral triangles at the center of the top of the GACS are fractal models for iterated function system characteristics. The Threshold Stabilization within the bottom of the GACS realizes the theory of games to allow stabilization of the sections where the explosives were detected by imaging in concrete sensors prior to attempts to detonate the explosives.

Using OPNET to Simulate an IP Network
Susan Spagnola Rutherford and Steven P. Gordon (Mitretek Systems)

Monday 1:30:00 PM 3:00:00 PM
Cases IV

Chair: Charles McLean (National Institute of Standards and Technology)

In this case study we demonstrate the value of simulation modeling in the healthcare architecture design of a new ambulatory surgery unit. Through simulation we were able to explore the relationship between the number of operating rooms and recovery beds, the number of cases and the phases of recovery, bed capacity, staffing, and waiting room seat requirements as it affected the design. This information allowed time for variations in the design to meet the client's needs before the project was built. Furthermore, we show the positive outcome of combining the experience of three professional disciplines: healthcare planning, architecture, and simulation modeling for nurses, doctors and patients of the new ambulatory surgery unit.

Using Simulation to Improve the Patient Pre-admission Testing Design Process
Susan O'Hara, RN MPH (O'Hara HealthCare Consultants), Mark Sullivan, AIA, NCARB (Mark Sullivan Architects) and Henry Bell (Smoky Mountain Simulation Services)

In this case study, we demonstrate how simulation modeling affects more than healthcare design; it is a tool of hospital-wide system process change with a focus on improving patient satisfaction through a better clinical process. Some of the improvements, based on existing systems, include decreased delays, better healthcare staff utilization and tailored scheduling. In this part 2 of 2 perioperative case studies(part 1: Ambulatory Surgery Unit), we maximized and combined the experiences of three distinct, yet complimentary professions: nursing/healthcare planning, architecture, and simulation engineering to create a better way to provide a clinical service in a newly designed healthcare space.

The Use of Simulation Modeling in Healthcare Planning
Susan O'Hara, RN, MPH (O'Hara HealthCare Consultants), Dave Krahl (Imagine That, Inc) and Mark Sullivan, AIA, NCARB (Mark Sullivan Architects)

Monday 3:30:00 PM 5:00:00 PM
Cases V

Chair: James Wilson (North Carolina State University)

A Pharmaco-economic Arena Model Comparing Pacemakers
Jörgen Möller, J. Jaime Caro, and Alexandra Ward (Caro Research Institute)

A model showing the pharmaco-economic effects of two different pacemakers in 1000 patients by twinning the patients once their characteristics are assigned. The model (in Arena) looks at both the "personal" effects and the aggregated economic implications.

Simulation Lays the Foundation for Artificial Intelligence in Surgery Center
Christian Baldwin (Engineering Innovations)

In general operating suites, there is a problem stemming from a misalignment of clinical staff and operating resources. The main objectives are to increase the throughput and associated revenues, increase patient safety, and decrease operational costs. Simulation can be used to model the current environment and experiment with different scheduling algorithms and resource assignments until an optimal method of operations is derived. If planned correctly, significant results can be achieved after the first iteration of implementation; however, additional gains in capacity are mathematically possible if you work to increase communications with artificial intelligence. Simulation is just one key component to the total AI engine. Further advances in the world of simulation and the relationship of simulation to other fields of science will provide not only the foundation for artificial intelligence but also the constructs for real-time management and proactive decision making for the healthcare industry for many years to come.

Tuesday 8:30:00 AM 10:00:00 AM
Cases VI

Chair: David Goldsman (Georgia Institute of Technology)

The purpose of this presentation to familiarize the participants with the joint USAF/1st AF/Joint Theater Air and Missile Defense Organization initiative entitled “Critical Area Air Defense (CAAD) Architecture Assessment” that began in 2002. The CAAD briefing is in two parts: (1) the discussion of the CAAD purpose, objectives, architecture, infrastructure; and (2) a description of the events themselves. CAAD is a joint government initiative. Since its inception in January 2002, USAF ESC has taken the technical lead, joined with First Air Force operational lead and Joint Theater and Missile Defense Organization (JTAMDO), to design, construct, implement, and apply a simulation based infrastructure to assess Homeland Air Defense (HLAD) architectures against cruise missiles and terrorist threats similar to the 911 attacks. Raytheon Integrated Defense Systems (IDS) provided the necessary support to ESC and 1st AF as the prime contractor and technical lead.

Multisensory Integration for Pilot Spatial Orientation
Ronald Small (Micro Analysis & Design, Inc.), Christopher Wickens (University of Illinois) and John Keller and Alia Oster (MA&D, Inc.)

Spatial disorientation (SD) is a normal human response to accelerations in flight. Its cost to the US military is over $300 million per year, with comparable costs to US civil aviation. Despite significantly increased research over the past decade, the rate of accidents caused by SD has not decreased. While the most recent research emphases have been on understanding the physiology of SD, the translation of the new knowledge into tools (e.g., training, displays, automation) that help pilots avoid SD and minimize its effects if it does happen, has not occurred. Our research goals were to apply multisensory countermeasures (CMs) to SD based on human sensory models and the pilot’s workload. We used SD mishap flight simulations to test our models and pilot SD Aiding System.

Spares Optimization for the Lockheed Martin Joint Strike Fighter
Jean-Paul Watson, Michael Collins, William Hart, Hai Le, Laura Swiler, and Bruce Thompson (Sandia National Laboratories)

Over the next few decades, the Lockheed Martin Joint Strike Fighter will supplant existing aircraft in various branches of US and NATO armed forces. A major deployment challenge is the identification of low-cost, high-performance sparing policies and resource allocations across geographically dispersed supply and repair chains. Through a joint effort, Sandia National Laboratories and Lockheed Martin have developed a detailed enterprise-level discrete events simulation of the complete operation. Optimization in the context of this simulator is exceptionally challenging: the number of free parameters numbers in the hundreds of millions, while individual simulations require hours of run-time. We will describe the architecture and algorithms forming the basis of a software system that leverages a careful balance of simulation output analysis, heuristics, mixed-integer programs, and local search to achieve low-cost, high-quality sparing policies and resource allocations for the full-scale distribution network.

Critical Area Air Defense (CAAD) Homeland Air Defense Architecture Assessment
James M. Gill (Computer Sciences Corporation), Tom Sadeghi (Raytheon Integrated Defense Systems), Steven Boe (1st Air Force), Hutch Davis (1st Air Force) and Kari Perez (ESC)

Tuesday 10:30:00 AM 12:00:00 PM
Cases VII

Chair: K. Preston White (University of Virginia)

Optimal Intruder Detection via a Multiobjective Evolutionary Algorithm
Keith Montgomery Sullivan (Systems Planning and Analysis, Inc.)

The current geopolitical environment has focused attention on protecting military and civilian installations from malicious trespassers. This work examined the detection of intruders attempting to slip through a secure area undetected. Aiming to minimize the number of undetected intruders, the defenders use a combination of fixed sensors, mobile robots, and a human reaction force to evaluate each intruder. The simultaneous objectives of maximizing the number of detected intruders and minimizing overall cost, plus the unknown interactions between the three elements of the defense and the unknown shape of the state space, led to the use of a multiobjective evolutionary algorithm. For different reaction force time delays and multiple intruder formations, data was collected on the percentage of targets assessed (and the corresponding system configuration), location of assessment, time to assessment, and reaction force detection range. Our customer used this data to assist with acquisition and personnel decisions.

Simulating Secured Environments
Lester Cano (Sandia National Labs) and Jonathan C Phillips (Rockwell Automation)

Abstract not available.

A Simulation Based Model for Default Risk in HEL Subordinate Securities
Kishore K Yalamanchili and David H Bai (State Street Research & Management)

The Home Equity Loan (HEL) ABS sector has grown to be a significant part of the ABS market. The securities offer attractive yields, particularly the subordinate bonds. The cash flows of subordinate bonds have a complicated pattern and are dependent on the status of certain performance tests. Investors in the securities typically perform default analysis of HEL securities using a scenario based approach. We propose a model for evaluating the default risk in subordinate HEL securities backed by sub-prime first lien loans. Our model is based on simulation of serious delinquencies and losses. We show that the current scenario based approach does not adequately capture the risk inherent in these bonds.

Tuesday 1:30:00 PM 3:00:00 PM
Cases VIII

Chair: Barry Nelson (Northwestern University)

The vision of the Space and Missile Center (SMC) of the Air Force is to achieve excellence in space system acquisitions by evolving as a process-centric operation. A team of experts from The Aerospace, SAIC, Sparta, and Northrup-Grumman are working under the SMC Directorate of System Acquisitions and the SMC Chief Engineer’s Office to help develop the concept of operation, strategic plans and the infrastructure required to implement managed processes. The dynamic model example presented in this simulation case study attempts to provide a quantitative perspective of the process management CONOPS and paradigm change by taking processes through their life cycle. In the near future, the model will be populated with all SMC processes. The model uses Extend v6 with operations research module as platform. The author of the model and the demonstration movie is a member of the SMC process team and a senior principal investigator at Sparta Inc.

Integrated Resource Planning Model
Richard Rinehart, Reid Kress, Tom Insalaco, Janice Atwater, and Jack Dixon (BWXT Y-12, LLC)

The Integrated Resource Planning Model (IRPM) is a decision support tool for dynamic rough-cut resource and capacity planning. The IRPM evaluates the impact of rapidly changing schedules, variable budgets, and a changing resource base including personnel, equipment, and facilities and their effects on schedule performance, projected cost, and resource utilization. While the IRPM has been developed with a powerful embedded Supply Chain simulation model as its solution engine, it employs an easy to use graphical user interface for the novice analyst. The key source of data for the model is program and project schedules and associated resource and material connectivity information. The IRPM has been demonstrated successfully at the Y-12 National Security Complex in Oak Ridge, Tennessee, on a large program and on a simple demonstration model of a bicycle refurbishment/manufacturing facility. This demonstration will highlight the simulation model, the user interface, and the intercommunications between the two.

SMC Process Management Dynamic Model
Moiz Baig (Sparta, Inc.)

Tuesday 3:30:00 PM 5:00:00 PM
Cases IX

Chair: Deborah Sadowski (Rockwell Software)

Improving Distribution Center Operations Through Simulation At Crutchfield Corporation
Paul G. Sherland (Rust Management Technology, LLC) and David Luria (Crutchfield Corporation)

Crutchfield Corporation is a major consumer electronics catalog and Internet retailer with a reputation for outstanding customer service and support. The company is applying Six Sigma techniques to improve distribution center operations; however, wide variation in order arrival rates and the types of orders make workflow analyses difficult. Rust Management Technology, LLC, proposed developing a process simulation of the distribution center order handling workflow that could be used to study the impacts of changing order arrival patterns, labor allocations, and worker efficiencies. Working as a team, Crutchfield Corporation and Rust Management Technology used Extend and ExpertFit to develop and validate a distribution center process simulation. Using insights developed through the simulation, Crutchfield Corporation estimates that it will save 15 to 20 times the cost of the project in the first year after deployment.

Artisan Bread Manufacturing Simulation
Alison R. Kulp (Productivity Management Consultants)

A large artisan bread company integrated automated elements such as conveyors, laser sensors, weighing devices and slicing machines with manual components such as hand kneading, loaf forming and scoring in a flexible manufacturing line capable of producing over 25 different types of bread in various shapes and sizes. A simulation of this system successfully identified bottlenecks and allowed additional capacity to be obtained. As a result, the company was able to postpone the purchase of a new manufacturing line by six months through the discovery of alternative ways to increase the output of the system.

Simulation Optimization Decision Support System for Ship Panel Line Operations
Charles M. LaRue (Northrop Grumman Ship Systems - Ingalls Operations), Allen G Greenwood (Mississippi State University), Travis W Hill and Jeffery Wilson Miller (Mississippi State University - Center for Advanced Vehicular Systems) and Clayton Thomas Walden (Mississipi State University - Center for Advanced Vehicular Systems )

This university-industry project developed, in addition to discrete-event simulation models of the bottleneck shop in two major shipyards, a decision support system (DSS) that integrates the simulation models and an evolutionary-strategy-based optimizer. The system determines the optimal sequence for producing ship panels (each panel is unique) based on each panel’s physical characteristics and work content, resource availability, earliest start and latest finish dates, operating policies, and business rules. The quality of each alternative sequence is assessed through detailed simulation models of the shops. The DSS can use models developed in either ProModel or QUEST. The system is used by planners to assess shipyard capacity and by shop-floor supervisors to better allocate resources. The successful results demonstrate the capability to effectively imbed and hide complexities associated with sophisticated simulation and optimization technologies from non-technical users, yet enable the users to obtain the benefits of using the technologies.

Wednesday 8:30:00 AM 10:00:00 AM
Cases X

Chair: Susan Sanchez (Naval Postgraduate School)

Planners of batch chemical processes face a wide range of challenges when scheduling production to ensure adequate product to fill promised orders. Today, solving these scheduling problems using rigorous optimization methods is feasible in the standard business time frame, ie, fast enough that the user perceives that it makes their job easier. This case-study will present multi-product batch optimal scheduling using Resource-Task Networks (RTN). Although this deterministic approach is successful, the stochastic nature of the process inevitably leads to rescheduling. The structure of the RTN allows for nearly one-to-one mapping into a discrete-event simulation (DES). The DES allows us to refine the grid size and input parameters for the optimization. The DES also allows for quantitative comparisons between "by-hand" schedules and the proposed optimal schedules from the RTN. This combined approach of discrete-event simulation and deterministic optimization has provided rapid improvements in the scheduling technology.

Applying Simulation to the sizing of PETROBRAS’ oil refinery tank farms

The simulation model uses scenarios to support decisions on the sizing of a refinery tank farm. It analyzes the impact of new enterprises, demand variations and detects operational bottlenecks. It is made up of a mass balance associated to a regime consisting of the entry of feedstocks, exit of outlet products and handling operations subject to prioritization rules. The arrival of crude oils, serving the oil products' consumer market and operational failures are the random events considered. Its great flexibility allows many variations to the process elements, as well as to the quantity and volume of the tanks. Studies carried out have shown adherence and indicated necessary process adjustments. General routines will be used to replicate the model, consolidating it as an indispensable instrument in the Company's refineries investment analyses. At that stage, the models will have a user friendly interface and will access corporate databases to facilitate their calibration.

Ensuring the Production Facility Achieves its Required Capacity and Exceeds It
Joseph Weiss (Alfa Laval Biokinetics)

A biotech facility was originally designed without simulation. The use of simulation to confirm the design identified that the plant would be unable to meet the desired production requirements. By modeling and evaluating multiple scenarios, the most cost-effective solution was identified that enabled the facility to achieve its production goals. In order to ensure an effective solution for the entire facility, all processes, support systems, and critical utilities were modeled and evaluated. When production forecasts showed a significantly increased demand, the simulation model was revisited. The production of the facility was increased 50% with only a 7% increase in capital cost using constraint sequence evaluation.

Combining Simulation with RTN Optimization for Multi-product Chemical Processes
Scott J Bury, Naoko Akiya, and Jeff Ferrio (The Dow Chemical Company)

Wednesday 10:30:00 AM 12:00:00 PM
Cases XI

Chair: Ernest Page (U.S. Army Model and Simulation Office)

The DuPont Company has chosen the Six Sigma methodology as "the way we do business". Six Sigma's demand for quantifiable bottom line results is an excellent business discipline. At the same time, for many multi-stage production and supply chain operations, it is difficult to quantify the impact of a given project on overall earnings. Simulation is used in DuPont to help select projects which have high potential for increasing earnings. When the projects meet their immediate goals, simulation is used to quantify the contribution of that project to overall earnings improvement. A case study of a large multi-step chemical operation will be reviewed.

Strategic Analysis, Modeling and Simulation (SAMS) of logistics as part of a planned vertical mega shaft system
Martin M Franklin (Ceenex LC) and Jan-Diedeleff Gertenbach (Ceenex Pty Ltd)

Impala Platinum Limited mining operations focus on one of the largest known deposits of platinum group metals (PGMs) in the world, namely the Bushveld Complex in South Africa. Impala is investigating the feasibility of a planned expansion of their operations, specifically a new vertical mega shaft system incorporating innovative mining methods. Dynamic simulation modeling toolkits were developed in an object orientated environment for rapid application during the feasibility analysis of the logistics involved with the shaft system. The logistic models address the complete mining process, including materials handling operations underground and on surface (therefore incorporating hoisting, rail transport, truck haulage, conveyors, etc.). SAMS provide Impala with valuable decision making and what if analysis. This presentation provides an overview of SAMS applied to Impala Platinum requirements and demonstration of the modeling toolkits.

Applying Specialized Simulators for Kanban and Flow Optimization (KanbanSIM) and Powertrain Manufacturing (PTMSim)
Samarn Chantaravarapan (Production Modeling Corporation (PMC)) and Onur Ulgen (Production Modeling Corporation and University of Michigan-Dearborn)

We present two specialized simulators: KanbanSIM and PTMSim. KanbanSIM is a simulator that optimizes the Card/Call (Electronic) Kanbans in the line while minimizing the material handling flow in the system. It captures the production variability at assembly stations, shows the effects of material handling equipment scheduling, etc. Its application is an Automotive Final Assembly Line with about 300 stations and 3000 Kanban types. The selection and scheduling of the material handling equipment and the station inventory sizes were based on the analysis using KanbanSIM. PTMSim, Power Train Manufacturing Simulator, is a powerful tool for automotive industries. PTMSim provides users predefined and parameterized manufacturing objects that are typical to such manufacturing environments. PTMSim reduce the developing time by 60-70 percent and model verification and validation time by 50 percent. Three-dimensional animation encourages a better vision and the model can be built on top of the AutoCAD layout for precise physical dimensions.

Using Simulation to Select and Evaluate Potential Six Sigma Projects
J Bennett Foster (DuPont Company)