i-PICS: Integration of Advance Demand Information with Pull-Type Production Control Systems

产品种类的增加和供应链之间高级需求信息(ADI)的共享正在重新定义工厂运营中的基本权衡。为了获得工厂效率的巨大提高，供应链上可用的信息必须与生产库存控制系统(PICS)集成，生产库存控制系统确定工厂内不同产品的生产计划和库存数量。这些集成的PIC(i-PIC)可以对工厂效率(生产流程时间、产能利用率和交付绩效)产生重大影响。本研究旨在建立一种新的排队网络模型，用于评价具有产品多样性和ADI的制造系统中PICS的性能。这些模型将被用于(I)量化ADI与拉式PICS整合所获得的运营改进，(Ii)研究ADI质量对整合PICS潜在效益的影响，以及(Iii)在博弈论环境下获得供应链中的信息共享策略(合作与竞争)的洞察。研究将导致新的可扩展且计算高效的方法来分析一类重要的具有同步站的封闭排队网络。该方法基于节点分解和流量过程近似，不需要对概率分布的限制性假设，从而使它们能够用于分析制造系统的现实模型。这项研究将构成多篇博士论文的基础，并将为伦斯勒理工学院的本科生提供研究机会。将设计一门新的研究生/本科课程，并开发额外的教学模块，以加强现有课程。这项研究还将为当地行业的学生提供基于团队的项目体验。行业数据和实施将验证这项研究，其见解将与纽约经济增长中心及其行业附属机构合作传播。

The increase in product variety and sharing of advanced demand information (ADI) across supply chains are re-defining the fundamental tradeoffs in factory operations. To obtain quantum improvements in factory efficiencies, the information available across supply chains must be integrated with the production inventory control systems (PICS) that determine the production schedules and inventory quantities for the different products within the factory. These integrated PICS (i-PICS) can have a significant impact on factory efficiency (production flow times, capacity utilizations and delivery performance). This research is to develop new queuing network models for performance evaluation of PICS in manufacturing systems with product variety and ADI. These models will be used to (i) quantify the operational improvements obtained from the integration of ADI with pull-type PICS, (ii) investigate the impact of quality of ADI on potential benefits of integrated PICS, and (iii) obtain insights into information sharing strategies (collaboration versus competition) in supply chains in a game theoretic setting.The research will lead to new scalable and computationally efficient approaches to analyze an important class of closed queuing networks with synchronization stations. The methodology is based on node decomposition and traffic process approximations that do not require the restrictive assumptions on probabilistic distributions, thereby enabling their use to analyze realistic models of manufacturing systems. This research will form the basis of the multiple doctoral theses and will provide undergraduate student research opportunities at Rensselaer Polytechnic Institute. A new graduate/undergraduate course will be designed and additional teaching modules will be developed to enhance existing curriculum. This research will also provide team-based project experience to students at local industry. Industry data and implementations will validate this research and the insights will be disseminated in collaboration with the Center for Economic Growth in New York and its industry affiliates.