Simulation platform for anticipative plant-level maintenance decision support system
提出一个预测性工厂级维护决策支持系统，基于设备瓶颈排序指导纠正性和预防性维护优先级，以提升日产量，并用汽车车身车间的真实数据验证其优于其他方法。
Global competition and increasing customer expectations are forcing automobile manufacturers to improve their operations. Maintenance, being one of the most critical components in many industries, has a direct impact on the improvement of the overall production performance. In this paper, we introduce an anticipative plant-level maintenance decision support system (APMDSS) that provides guidance on corrective and preventive maintenance priorities based on the equipment bottleneck ranks with the objective of improving daily plant throughput. APMDSS anticipates the plant dynamics (i.e. bottlenecks, hourly buffer levels and likelihood of machine breakdowns) for upcoming shifts using starting state information of the production shift (e.g. equipment maintenance history, operational status of machines, buffer levels and scheduled production model mix). We also evaluate the performance of APMDSS using real data from an automotive body shop experiencing routine throughput difficulties due to frequent machine breakdowns. The results are compared with other methods from the literature and found to be superior in many settings.