GOALI: Incorporating Nonlinear Phenomena into Production Planning Models

这项“与工业界学术联络资助机会”（GOALI）奖为整合生产计划活动的两个关键方面的研究提供资金，这些活动迄今为止是使用非常不同的模型分别进行的：使用数学规划模型对工厂工作的时间阶段进行规划，并建立安全库存水平，以保证客户服务的期望水平，同时在生产资源的能力限制下保持可行。连接这两个操作的方面是交货时间，即从产品进入工厂到成品出现所需的时间。众所周知，由于对资源利用率的影响，工作发布的数量和时间决定了提前期。另一方面，安全库存水平的建立需要了解生产设施补充库存的前置时间。从历史上看，这两个问题一直是孤立地研究的，但是最近在普渡大学，nsf支持的工作第一次允许开发发布计划模型，该模型可以正确地捕获发布对交付时间的影响。这提供了一种在集成模型中将安全库存确定与放行计划联系起来的机制。如果成功，这项研究将导致新的供应链规划工具的发展，这将允许涉及复杂制造设施的大型供应链在大幅降低库存水平的情况下运作，大幅降低成本，而不会对服务水平产生不利影响。它还将有助于更好地理解两种研究流之间的联系，这两种研究流在过去被视为使用非常不同的数学工具的不同领域。

This Grant Opportunity for Academic Liaison with Industry (GOALI) award provides funding for research on means of integrating two critical aspects of production planning activity that have hitherto been performed separately using very different models: the planning of the time-phased release of work into factories using mathematical programming models, and the establishment of safety stock levels that guarantee a desired level of customer service while remaining feasible with respect to capacity constraints on the production resources. The aspect linking these two operations is the lead time, the time required for work released into the plant to emerge as finished product. It is well known that the amount and timing of work release determines lead times, due to its effect on resource utilization. The establishment of safety stock levels, on the other hand, requires knowledge of the lead time of the production facility replenishing that inventory. Historically the two problems have been pursued in isolation, but recent NSF-supported work at Purdue has for the first time allowed the development of release planning models that capture the effects of releases on lead times correctly. This provides a mechanism for linking the safety stock determination with the release planning in an integrated model. If successful, this research will lead to the development of new supply chain planning tools that will allow large supply chains involving complex manufacturing facilities to operate with substantially lowered inventory levels, substantially reducing costs without adversely affecting service levels. It will also contribute to a better understanding of the linkages between two research streams that have been treated as distinct areas with very different mathematical tools in the past.