SBIR Phase I: A Bottleneck Dynamics Decision Support System for Scheduling and Control

点击翻译按钮获取中文摘要

9560046 Morton This Small Business Innovation Research (SBIR) Phase I project develops a new simulation-oriented methodology for fast, accurate, and intuitive scheduling in manufacturing operations. Under escalating pressures of zero inventories, time based competition, and make-to-order requirements, finite scheduling is emerging in industrial practice. However, current methods are simple simulation approached with manual control or else rather complex and expensive automatic schedulers that allow little interaction with the user. This methodology, called bottleneck dynamics, is expected to allow a full range of manual, automatic, or mixed control and work fast. Bottleneck dynamics is intuitive in many shops where individual activities do not dominate, i.e., the structure is close to convex (in the aggregate). This suggests that: (a) heuristic dual prices are meaningful, providing approximate sensitivity analysis, (b) such prices can be used to fashion a broad variety of heuristics, including transfer pricing between different levels and parts of the shop, and (c) heuristic prices can be improved by iteration (hill climbing). This approach to scheduling has been explored in computational studies on automatic sequencing for a variety of objectives. The SBIR Phase I will extend bottleneck dynamics to routing, parallel machines, and sequence dependent setups through laboratory testing. Phase II is projected to add a decision support interface and two-level scheduling and control, and to test a prototype of a commercial system. Bottleneck dynamics is expected to provide small manufacturers with planning and scheduling software tools in manual, automatic and mixed modes of use. It is expected to support them in fast, accurate, easy-to-implement-and-use finite capacity scheduling opereations at affodable costs.