WSC 2002 Final Abstracts

The Role of Simulation in Advanced Planning and Scheduling
Kenneth Musselman, Jean O'Reilly, and Steven Duket (Frontstep, Inc.)

Abstract:
The tasks of planning and scheduling in manufacturing have evolved from simplistic Material Requirements Planning systems to today’s sophisticated Advanced Planning and Scheduling systems. While planning is concerned with the long-range determination of what needs to be manufactured, typically over a relatively long time period, scheduling is the task of deciding how that manufacturing is to be accomplished, typically over a relatively short time period. Simulation is well suited to the scheduling task since it can handle as much detail as is necessary to capture the subtleties of the manufacturing process. It is desirable for a simulation-based scheduling function to be integrated with an Enterprise Resource Planning system, which maintains the system data suitable for driving a simulation of the current system load and thereby producing a feasible schedule. This paper describes such an integrated system and the role of simulation within it.

Promise and Problems of Simulation Technology in SCM Domain
Sam Bansal (SAP Americas Inc)

Abstract:
This paper begins by identifying the potential Promise of Simulation domain. It also provides a brief review of this domain and modeling methodologies as applied to supply chain optimization. Problems and solutions of this area are discussed forming the rationale behind most of the industrial practice of this author. As a result most of the deterministic Business Process Reengineering and Opportunity Assessment work that needs to be done resorts to the "a priori methods" Building the simulation models costs more time and effort than implementing an equivalent solution from SAP such as APO or any part thereof in the domain of Supply Chain Management and Optimization. Against this environment and e-Supply Chain Management as a domain of the focus, this paper describes the methodology of doing Business Cases with Case Studies to illustrate how the Supply Chain Opportunity Assessment through the Blue Printing process is carried out.

Using Simulation to Evaluate Buffer Adjustment Methods in Order Promising
Hank Grant and Scott Moses (University of Oklahoma) and Dave Goldsman (Georgia Institute of Technology)

Abstract:
Much literature exists for scheduling production, but there is little work on establishing the due dates that serve as the inputs to developing a production schedule. We call this Order Promising. This paper explores a simulation-based approach for evaluating methods for promising the delivery of orders based on dynamic buffer adjustment coupled with various methods to forecast the amount of buffer required. The primary objective of the paper is to frame the problem and suggest methods of analysis. Preliminary computational results are also presented.

ASAP Applications of Simulation Modeling in a Wafer Fab
Kishore Potti and Amit Gupta (Texas Instruments, Inc.)

Abstract:
The authors define 4 levels of complexity in simulation modeling. The ability of the models to predict bottlenecks in the fab, capability of the model to be used for strategic applications such as cycle time reduction, simulation of complex dispatch rules using the model, capability of the model to predict operational output of the wafer fab that is clean room outs by product by day. This paper presents the operational applications of the ASAP simulation model to provide WIP flush to the test probe area and the flush provided to planning for financial estimates and fab commitments. WIP flush is defined as a prediction of the fab output by device/technology by day. In general the ability to predict the shorter the time horizon the more difficult it is to predict the output accurately. Excursions are defined as any deviations in the normal processing of wafers such as unusually high particles being shown on Statistical Process Control charts etc.

Design, Development and Application of an Object Oriented Simulation Toolkit for Real-Time Semiconductor Manufacturing Scheduling
Chin Soon Chong (Gintic Institute of Manufacturing Technology) and Appa Iyer Sivakumar and Robert Gay (Nanyang Technological University)

Abstract:
Real-time scheduling of semiconductor manufacturing operations, semiconductor test operations in particular, is complicated due to the following factors; multi-head resources, multi-level hardware dependency, temperature and hardware criteria, dynamic determination of processing time and indexing time, batch processing and re-entrant flow. A first-of-its-kind, object oriented (OO), discrete event simulation (DES) toolkit, RTMSim++ for real-time simulation-based scheduling applications has been conceptualized, designed, developed to resolve real-time scheduling problems in manufacturing. This paper reviews the work done in the development of RTMSim++ toolkit, and a case study in a real-time scheduling application. The salient features of the toolkit includes flexibility to customize and extend its functionality, real time shop floor status data initialization, and capability for modeling complex resource and process relationships. In the case study, RTMSim++ has been customized to incorporate very company specific heuristic rules, with the objectives of improving delivery performance, equipment utilization and cycle time.

Using Simulation-Based Scheduling to Maximize Demand Fulfillment in a Semiconductor Assembly Facility
Juergen Potoradi and Ong Siong Boon (Infineon Technologies), Scott J. Mason (University of Arkansas) and John W. Fowler and Michele E. Pfund (Arizona State University)

Abstract:
This paper describes how a large number of products are scheduled to run in parallel on a pool of wire-bond machines to meet weekly demand. We seek to maximize demand fulfillment subject to system constraints. The schedule is generated by a simulation engine and used to control the machines at execution time and also to plan for the start of material. By using online data for equipment status and WIP availability, the schedule adapts to “unforeseen” changes on the shop floor after a simulation run. The frequently updated schedule redirects the line towards maximum demand fulfillment based on the latest status of the line.

Simulation based Multiobjective Schedule Optimization in Semiconductor Manufacturing
Amit K. Gupta and Appa Iyer Sivakumar (Nanyang Technological University)

Abstract:
In semiconductor manufacturing, it requires more than one objective such as cycle time, machine utilization and due date accuracy to be kept in focus simultaneously, while developing an effective scheduling. In this paper, a near optimal solution, which is not inferior to any other feasible solutions in terms of all objectives, is generated with a combination of the analytically optimal and simulation based scheduling approach. First, the job shop scheduling problem is modeled using the discrete event simulation approach and the problem is divided in to simulation clock based lot selection sub-problems. Then, at each decision instant in simulated time, a Pareto optimal lot is selected using the various techniques to deal with multiobjective optimization such as weighted aggregation approach, global criterion method, minimum deviation method, and compromise programming. An illustration shows how these techniques work effectively in solving the multiobjective scheduling problem using discrete event simulation.

Application of Simulation and Mean Value Analysis to a Repair Facility Model for Finding Optimal Staffing Levels
G. Boyer and A. N. Arnason (University of Manitoba)

Abstract:
Staffing problems arise in a wide range of applications including job shops, call centres, and hospital emergency departments. They are characterised by the need to allocate shift workers with varying skills to handle an arrival stream of tasks having different sub-task routings and (sub-task) skill requirements. The Manitoba Telecom Service Trouble Diagnosis and Repair System (TDRS) has 3 skill-levels of staff handling multiple types of faults occurring in telephone switching equipment. TDRS is a pure staffing problem having no equipment constraints: the only resource constraint is staff itself. The object of this study is to show how this can be modelled as an open network of queues with feedback and allowing for temporal and fault-class heterogeneity. Analytic mean value analysis then facilitates validation and selecting feasible staffing strategies for closer examination by simulation. The purpose of experiments using simulation is to find effective performance visualisations and "optimal" staffing allocations.

A Comparison of Three Optimization Methods for Scheduling Maintenance of High Cost, Long-Lived Capital Assets
Terry M. Helm, Steven W. Painter, and W. Robert Oakes (Los Alamos National Laboratory)

Abstract:
A range of minimization methods exist enabling planners to tackle tough scheduling problems. We compare three scheduling techniques representative of “old” or standard technologies, evolving technologies, and advanced technologies. The problem we address includes the complications of scheduling long-term upgrades and refurbishments essential to maintaining expensive capital assets. We concentrate on the costs of being able to do maintenance work. Using a standard technology as the baseline technique, Constraint Programming (CP) produces a 50-yr maintenance approach that is 31% less costly. Genetic Programming produces an approach that is 60% less costly.

A Simulation Model for Field Service with Condition-Based Maintenance
Yiqing Lin, Arthur Hsu, and Ravi Rajamani (United Technologies Research Center)

Abstract:
In this paper, we consider field service in which service providers are responsible for maintaining equipment performance. To do so, preventive maintenance work is usually required in addition to repairs. Field service managers are often faced with the conflicting objectives of maintaining a high level of equipment availability and keeping a low service cost. A condition-based maintenance (CBM) system can be used to achieve both goals by using equipment condition as a guide for taking maintenance actions. We developed a simulation model for field service with an integrated CBM system. A test case based on the field service operation of an elevator service provider has been built in a visual simulation environment to estimate the value of a CBM system.

Rolling Horizon Scheduling in Large Job Shops
Kristin A. Thoney, Jeffrey A. Joines, Padmanabhan Manninagarajan, and Thom J. Hodgson (North Carolina State University)

Abstract:
The Virtual Factory is a job shop scheduling tool that was developed at NC State. It has been shown to provide near-optimal solutions to industrial-sized problems in seconds through comparison to a computed lower bound. It is an iterative simulation-based procedure, whose objective is minimizing maximum lateness. Like many other job shop scheduling tools, the Virtual Factory has been evaluated primarily in a transient setting, even though a rolling horizon setting is more indicative of the situation in which scheduling algorithms are used in industry. Consequently, a rolling horizon procedure has been developed with which the Virtual Factory was tested. Experimental results indicate that the Virtual Factory also performs well under these circumstances.

Shop Floor Scheduling with Simulation based Proactive Decision Support
Amit K. Gupta (Nanyang Technological University), Appa Iyer Sivakumar (Nanyang Tecnological University) and Sumit Sarawgi (A.T. Kearney Ltd.)

Abstract:
This paper involves the study of a simulation based proactive decision support module for the shop-floor scheduling of the Plastic Processing Section at Bharti Telecom Limited, Gurgaon, India. The flow of material and information in this shop is highly complex as it involves multiple product parts, sequence dependent setup, molding machine specifications, mould restrictions etc. with a variety of scheduling and operational choices. The shop floor planning and scheduling decisions are being exercised manually and there is enough scope of using this simulation based tool to improve the shop-performance. In this work, efforts have been made to simulate the scheduling environment of this section. The performance of the shop depends on various parameters such as initial conditions of the machines, the load on the system, sequencing rule etc. Also the user requirements are so varied and situation dependent that it requires the use of simulation techniques to get a better schedule.

Process Accompanying Simulation – A General Approach for the Continuous Optimization of Manufacturing Schedules in Electronics Production
Sebastian Werner and Gerald Weigert (Dresden University of Technology/IET)

Abstract:
The paper will present the successful realization of simulation based manufacturing planning and control in electronics production. The principle has been implemented in various applications as supplementary decision aid in connection with ERP systems. A theoretical model of virtual and real time explains relations between simulation and production planning. In all cases the discrete event simulator ROSI has been used, which provides comfortable functions for manufacturing models. Scenarios are being optimized by metaheuristic methods like Genetic Algorithms. As real production and simulation are producing data in the same logical format, a comparison technique is being proposed that enables the observation of errors between reality and simulation, where simulation can be retriggered when the error is rising too high. In addition the adaptation of parameters is derived of this comparison. Examples clarify all contributions.

Simulation Optimization for Process Scheduling through Simulated Annealing
Alex Cave, Saeid Nahavandi, and Abbas Kouzani (Deakin University)

Abstract:
A reinforcement learning agent has been developed to determine optimal operating policies in a multi-part serial line. The agent interacts with a discrete event simulation model of a stochastic production facility. This study identifies issues important to the simulation developer who wishes to optimise a complex simulation or develop a robust operating policy. Critical parameters pertinent to 'tuning' an agent quickly and enabling it to rapidly learn the system were investigated.

Simulation of JIT Performance in a Printing Shop
Ben M. Patterson, Mustafa Ozbayrak, and Theopisti Papadopoulou (Brunel University)

Abstract:
A medium sized UK based academic publishers own a subsidiary printing business. Presently the Academic Printers (AP) is experiencing productions line flow problems reducing the efficiency of the operation. Most of the problems are generated by the imbalanced workflow through the system. By implementing a JIT production planning system it is hoped that some of the production problems can be resolved. Using the simulation software a model was created to investigate the performance of the AP under a variety of operating conditions. Results showed that operating the system with JIT control would not produce economic performance improvements due to constraints applied by the printing process.

Simulation Study of Dreyer Urgent Care Facility
Boon Aik Tan, Aldas Gubaras, and Nipa Phojanamongkolkij (Northern Illinois University)

Abstract:
The health service center is considering modification of the current doctor schedule to decrease the average time patients spend in the facility. A new schedule, which on average has an extra doctor added for each time slot, has been suggested by the preliminary queueing analysis. It is anticipated that the schedule will improve the system performance but has yet to be verified by a discrete event simulation model. The shifting bottleneck issue has to be considered as well. Increasing the resource capacity of the current bottleneck (believed to be the doctor resource) might trigger a new bottleneck and could possibly worsen the current situation. In this study, the current and proposed doctor schedules are tested on the simulation model created using SIMAN language and simulated on Arena. The output from the simulation just verifies the preliminary queueing analysis that the proposed schedule decreases the average time patients spend in the facility. Sensitivity analysis is performed as well, and the next bottleneck resource is identified.