CAREER: Analysis of Multi-dimensional Coordination Problems in Service Parts Logistics Systems

该学院早期职业发展（CAREER）奖为服务零件物流（SPL）系统中协调问题的模型和算法的开发提供资金。SPL系统面临的挑战是协调跨业务单位、时间范围和地理位置的决策，同时实现高质量的全系统解决方案。研究首先将开发一个统一的框架，集成物流网络设计和库存管理，并探讨基于分解的解决方法制定的数学模型的相关问题。物流网络设计子模型涉及多梯队、多产品和基于时间的服务覆盖限制。库存子模型涉及跨多个设施的库存共享和基于响应时间的服务级别。由于真实的SPL系统中存在的问题的规模，本研究还将研究模型的不同粒度级别，并确定计算和解决方案精度之间的权衡。研究将最终解决机制设计和激励问题，内在地存在于协调问题：如何确保决策实体有必要的激励来协调他们的决策？该奖项的密切相关的教育计划包括现代SPL系统的教学案例的开发，以及供应链协调问题的数学建模课程的开发。如果成功，这项研究的结果将导致SPL问题的解决方案，无论是在规模和范围，共同面临的工业，政府和军队。这些模型及其结果将与SPL的合作伙伴一起进行验证和核实。这些研究将有助于为具有特殊结构的大规模数学模型及其在真实的问题中的应用寻找可扩展的求解方法。这些结果还将有助于使各个决策实体的目标与全系统的目标保持一致。在一个统一的框架中同时分析这些问题将有助于理解这些因素（不同层次的协调和激励问题）如何在分布式物流系统中相互影响。另一个贡献将是卧底协调机制和基于分解的优化技术之间的关系。教育部分应加强应用运筹学、供应链管理和物流课程的教学。

This Faculty Early Career Development (CAREER) award provides funding for the development of models and algorithms for coordination problems in Service Parts Logistics (SPL) systems. The challenge in SPL systems is to coordinate decisions across business units, time horizons, and geographical locations at the same time towards high-quality system-wide solutions. The research first will develop a unified framework for integrated logistics network design and inventory management and explore decomposition-based solution methodologies for the associated problems formulated as mathematical models. The logistics network design submodel involves multiple echelons, multiple products and time-based service coverage restrictions. The inventory submodel involves inventory sharing across multiple facilities and response-time based service levels. Motivated by the scale of the problems existing in real SPL systems, the research will also investigate different levels of granularity of the models, and identify the trade-off between computation and solution accuracy. The research will finally address the mechanism design and incentive issues that inherently exist in coordination problems: How does one make sure that the decision entities have the necessary incentives to coordinate their decisions? The closely related educational plan of the award involves the development of a teaching case on modern SPL systems, and the development of a course on mathematical modeling of supply chain coordination problems.If successful, the results of this research will lead to the solution of problems in SPL, both in size and scope, commonly faced by industry, the government and the military. The models and their results will be validated and verified with partners in SPL. The investigations should contribute in finding scalable solution methodologies for large-scale mathematical models with special structure and their application to real problems. The results will also provide insights in aligning the individual decision entities' goals with the system-wide goals. Analyzing these issues simultaneously in a unified framework will improve the understanding of how these factors (different levels of coordination and incentive issues) affect each other in distributed logistics systems. Another contribution will be to undercover the relationship between coordination mechanisms and decomposition-based optimization techniques. The educational component should enhance teaching in applied operations research, supply chain management, and logistics courses.