INTEGRATED PRODUCTION AND MAINTENANCE DECISION-MAKING UNDER EQUIPMENT DEGRADATION: IMPLICATIONS FOR OPERATIONAL PERFORMANCE

Abstract

Co-ordinating production and maintenance activities is essential in manufacturing systems but more so under a state of machine deterioration. Previous studies emphasized that additions between such decisions are essential in minimizing cost and maximizing productivity besides leading to an improved reliability of the system (Gharbi & Kenné, 2005; Nourelfath & Châtelet, 2012). Yet, most methods either ignore for the complexity of degradation dynamics, or do not consider machine deterioration as a random process. This paper formulates a Markov Decision Process (MDP) model to solve the integrated production-maintenance scheduling with degradation uncertainty. The goal is to optimize the tradeoff between production throughput and preventative maintenance intervention, while minimizing downtime and associated longterm costs. MDP facilitates explicit modeling of the probabilistic state transitions between machine health statuses which in turn supports making robust decisions under uncertainty (Puterman, 2014; Cassady & Kutanoglu, 2003). Results indicate that the proposed model is far superior to conventional periodic maintenance, static schedule approaches in terms of imminent total cost and system availability. The results indicate that such MDP frameworks can effectively be adapted to real-world dynamic production environments, where uncertainty is a significant factor (Bai & Yun, 2021). For practitioners, managerial practices are given that may be adopted by firms wishing to improve operational resilience and increase the life of equipment via data driven integrated scheduling strategies.

Keywords: Production scheduling, preventive maintenance, machine degradation, Markov Decision Process, stochastic optimization, system reliability.