Boon Ping Gan

Distributed supply chain simulation across enterprise boundaries

The effective practice of supply chain management (SCM) is crucial to improve corporations competitive advantage. Many corporations have built simulation models to facilitate the application of simulation in designing, evaluating, and optimizing their supply chain. Traditionally, a supply chain involves only a single enterprise with multiple facilities and distribution centers. Hence, sharing of detailed simulation models is not a problem in this scenario. But in recent years, the scope of SCM has evolved to cross the enterprise boundaries. Applying simulation in designing, evaluating, and optimizing the supply chain becomes more difficult since the participating corporations might not be willing to share their simulation models with partners. Distributed simulation techniques are presented as an enabling technology that allows corporations to construct a cross enterprise simulation while hiding model details within the enterprise. This can be realized by either building the simulation on top of the Runtime Infrastructure of the High Level Architecture or building the simulation on top of a customized distributed discrete event simulation protocol. These alternative approaches are compared in terms of their performance and interoperability. The comparison of the performance is done through a benchmarking test of a semiconductor supply chain model.

Distributed simulation with incorporated APS procedures for high-fidelity supply chain optimization

Tactical and operational planning for manufacturing enterprises are more important today than ever before as their supply chains span the globe. Two state-of-the-art technologies that are critical to success are discrete event simulation and advanced planning and scheduling. They are commonly applied in designing and executing operations at each site within the supply chain. However, as supply chains become leaner and more responsive, operational constraints and stochastic influences within the manufacturing sites and the logistics network require a combination of both technologies applied to the entire supply chain. This paper describes a novel framework for advanced distributed simulation with integrated APS procedures for collaborative supply chain optimization. The framework can be used for fast optimization of both planning procedures and execution policies and also provides a base for easy implementation of simulation results. A prototype of a distributed semiconductor supply chain simulation has been developed and is currently being refined.

Load balancing for conservative simulation on shared memory multiprocessor systems

Load balancing is a crucial factor in achieving good performance for parallel discrete event simulations. We present a load balancing scheme that combines both static partitioning and dynamic load balancing. The static partitioning scheme maps simulation objects to logical processes before simulation starts while the dynamic load balancing scheme attempts to balance the load during runtime. The static scheme involves two steps. First, the simulation objects that contribute to small lookahead are merged together by using a merging algorithm. Then a partitioning algorithm is applied. The merging is needed to ensure a consistent performance for our dynamic scheme. Our dynamic scheme is tailor-made for an asynchronous simulation protocol that does not rely on null messages. The performance study on a supply chain simulation shows that the partitioning algorithm and dynamic load balancing are important in achieving good performance.

Distributed Supply Chain Simulation as a Decision Support Tool for the Semiconductor Industry

The need for better understanding, control, and optimization of supply chains is being recognized more than ever in the new economy. Simulation holds a great potential in portraying the dynamic evolution of supply chains and providing appropriate decision support to address challenges arising from high variability and stochastic uncertainty. Realizing high-fidelity supply chain simulation will require integration of individual supply chain component simulation models and planning systems, shielding to prevent sensitive data from being shared indiscriminately, and even the geographical distribution of the supply chain component models. The authors discuss various conceptual and technical issues that have been successfully addressed to realize a prototype of distributed semiconductor supply chain simulation as well as implementation approaches that can be pursued. The prototype emulates a semiconductor supply chain consisting of two wafer fabs, an assembly and test facility, a distribution center, a warehouse, a supply chain planning module, a logistics provider, and customers.

Optimistic Synchronization in HLA-Based Distributed Simulation

With the wide use of commercial off-the-shelf (COTS) simulation packages and the advent of the High Level Architecture (HLA) standard, it is desirable to build distributed simulations by linking various types of simulation models developed using best-fit COTS packages. While almost all current work on integrating COTS packages and the HLA is based on conservative synchronization, it is worthwhile to investigate the optimistic synchronization approach. The optimistic approach can exploit parallelism and achieve promising performance in situations where causality errors may occur but in fact seldom occur. The authors introduce a rollback controller using a middleware approach to handle the complex rollback procedure on behalf of the simulation model. A new time advance algorithm is proposed that can fully use the benefits of optimistic synchronization. The article also describes a scalability study showing the experimental results for the two synchronization approaches as the number of simulation components increases.

Distributed simulation and industry: potentials and pitfalls

We present the views of five researchers and practitioners of distributed simulation. Collectively we attempt to address what the implications of distributed simulation are for industry. It is hoped that the views contained herein, and the presentations made by the panelists at the 2002 Winter Simulation Conference will raise awareness and stimulate further discussion on the application of distributed simulation methods and technology in an area that is yet to benefit from the arguable economic benefits that this technique promises.

Analysis of a borderless fab using interoperating AutoSched AP models

Semiconductor front-end manufacturing is a complex process involving a large number of fabrication steps that require capital-intensive equipment. To address shortages that can arise in individual wafer fabs in situations such as machine breakdowns or unexpected surge in demand for some products, the concept of a borderless fab is presented in this paper. In a borderless fab, multiple wafer fabs pool their capacity together, allowing movement of partially completed wafer lots from one fab to another. This enables the redistribution of workload to other similar resources in alternative fabs to maintain the targeted production cycle time. For the purpose of this study, two wafer fabs within close proximity and with similar processing capabilities were modeled using the AutoSched AP (ASAP) simulation package. Instead of creating one single large model, each wafer fab is modeled as an individual ASAP model. Simulation is executed on two computers interconnected by a local area network. The High Level Architecture (HLA) standard is adopted to enable the distributed execution. A novel time synchronization algorithm is proposed that is approximately ten times more efficient compared to conventional algorithms. The interoperating ASAP models were then used to investigate the effect of different lot batching sizes on lot transfer frequency and the average production cycle time in a borderless fab scenario. The experimental results show how an optimal operating point for lot batching size can be obtained from the intersection point of normalized contradictory performance indices. The operating point yields a balance between the number of lot movements between fabs and average cycle time.

Interoperating autosched AP using the high level architecture

The high Level Architecture (HLA) is an IEEE standard for interoperating simulation federates. In this paper, we describe a set of requirements that simulation packages need to satisfy in order to be made interoperable using the HLA standard. AutoSched AP, a commercial off-the-shelf simulation package (CSP) which is widely used in the semiconductor industry, was used as a case study for this interoperation exercise. We demonstrated that a straightforward customization of the CSP through a middleware that provides standard functions for interoperation may not provide a satisfactory solution. A specially optimized time synchronization mechanism needs to be installed to ensure good execution efficiency. Experimental results using a Borderless Fab model that comprises of two factory models show that an optimized time synchronization mechanism results in an execution time that is ten times better than a straightforward application of the HLA runtime infrastructures time synchronization mechanism.

Optimistic synchronization in HLA based distributed simulation

With the wide use of commercial off-the-shelf (COTS) simulation packages and the advent of the high level architecture (HLA) standard which supports interoperability and reusability, it is desirable to build distributed simulations by linking various types of simulation models developed using best-fit COTS packages. While almost all current work on integrating COTS packages and the HLA is based on conservative synchronization, it is worthwhile to investigate the optimistic synchronization approach. The optimistic approach can exploit parallelism and achieve promising performance in situations where causality errors may occur but in fact seldom occur. In our paper, we introduce a rollback controller using a middleware approach to handle the complex rollback procedure on behalf of the simulation model. To fully utilize the benefits of optimistic synchronization, we also introduce a novel time advance algorithm using the services provided by the HLA. A comparison of performance between the conservative and optimistic synchronization approaches based on a typical reference model is also provided.

Making use of prognostics health management information for aerospace spare components logistics network optimisation

Although research has evolved significantly over the last decade, there are still a large number of Grand Challenges confronting modelling, model deployment, and model-based decision making of large-scale complex Discrete Event Logistics Systems (DELS) to be tackled, as identified and reviewed during a Dagstuhl workshop in March 2010. This paper illustrates how several of these challenges are already being addressed, based on a series of case studies from the Aerospace Spare Components Logistics domain, where consolidated operational Prognostics and Health Management (PHM) information can be used for tactical planning and optimisation of spare components logistics networks. In this setting, the growing potential of PHM technology to facilitate the maintenance and support of commercial and military aircraft emphasises the need for tools to determine the impacts and benefits of a PHM system. To achieve this, the prognostics parameters and related logistics policies were identified, modelled, and subsequently incorporated into a simulation-based decision support framework.