DDDAS-SMRP: Dynamic Real-Time Order Promising and Fulfillment for Global Make-to-Order Supply Chains

本研究项目将研究全球订单制造供应链中与接收、接受、处理和履行客户订单相关的业务流程。它将研究这些过程背后的数据库和数据流、业务决策过程以及这些过程相互作用的方式。将开发模拟和优化模型，以监测输入数据的实时流。监控这些流的算法必须确定预测何时应该改变，并在适当的情况下计算这种变化。研究还将进行马尔可夫决策过程（mdp）的发展，动态适应传感器数据流。除了以MDP研究为基础之外，本主题还将考虑使用混合整数规划模型来设置MDP参数。研究结果将为改进按订单制造供应链的运营控制提供依据。它将为“始终在线”控制器的开发提供理论基础，该控制器可以持续监控供应链数据源并预测关键业务流程的未来状态。这项研究将支持在正常操作和事件响应下改进决策。支持的正常运营决策包括产品定位决策、订单时序和验收决策以及生产调度决策。事件响应模式将用于制定应对重大破坏性事件的策略。这些事件可能包括主要供应货物运输的失败或延迟，工厂的重大故障或接受了一份取代现有订单承诺的大订单。这些能力将大大提高客户服务水平和使用它们的供应链的盈利能力。

This research project will study the business processes related to receiving, accepting, processing and fulfilling customer orders in global make-to-order supply chains. It will investigate the databases and data flows underlying these processes, the business decision processes and the manner in which these interact. Simulation and optimization models will be developed that monitor real-time streams of input data. Algorithms that monitor these streams must determine when predictions should change and should calculate such changes if appropriate. Research will also be carried out into the development Markov Decision Processes (MDPs) that dynamically adapt to sensor data streams. In addition to building on MDP research, this topic will consider the use of Mixed Integer Programming models for setting MDP parameters. The results of this research will provide the basis for improved operational control of make-to-order supply chains. It will provide the theory underlying the development of "always on" controllers that continuously monitor supply chain data sources and project the future status of key business processes. This research will support improved decision making under both normal operations and incident response. The normal operations decisions supported could include product positioning decisions, order timing and acceptance decisions and production scheduling decisions. The incident response mode will be used to develop a strategy for reacting to major disruptive events. Such events might include the failure or delay of a major supply shipment, a major breakdown in a factory or acceptance of a large order that preempts existing order commitments. These capabilities should substantially increase the level of customer service and the profitability for supply chains that employ them.