Event-Time Models for Supply Chain Scheduling

Summary This study presents a modeling paradigm for scheduling problems in supply chains. The constituents of a supply chain need to cooperate, rather than compete, in order to achieve maximum respective benefits. To analyze this, it is essential to have a concise but comprehensive formulation of this problem. This formulation, in addition to being computationally viable, must account for special characteristics of supply chains. It is argued that lot streaming provides an appropriate paradigm for scheduling problems in supply chains. It is shown that event-time models provide a general formulation for lot streaming problems. Furthermore, logic-based modeling framework of constraint programming makes it possible to handle special requirements of these models. This is illustrated by a small example. It is also shown that the fundamental results for single-job lot streaming problems can be systematically obtained as special cases of this generalized formulation. This demonstrates that event-time formulation accurately models all of the features of the problem; and that starting with a concise but comprehensive formulation, solutions for special cases of the problem can be obtained systematically. To that end, a number of additional special case formulations of single-job lot streaming problem is presented. An appendix presents a schema for lot streaming problem classifications.