Distributed maintenance planning in manufacturing industries

Abstract

The combination of sensors and computing infrastructure is becoming increasingly pervasive on the industry shop-floor. Such developments are enabling the automation of more and more industrial practices, and are driving the need to replace conventional planning techniques with schemes that can utilize the capabilities of Cyber-Physical Systems (CPS) and Industrial Internet of Things (IIoT). The future is a place where intelligence is endowed to every entity on the shop floor, and to realize this vision, it is necessary to develop new schemes that can unlock the potential of decentralized data observation and decision-making. Maintenance planning is one such decision-making activity that has evolved over the years to make production more efficient by reducing unplanned downtime and improving product quality. In this work, a distributed algorithm is developed that performs intelligent maintenance planning for identical parallel multi-component machines in a job-shop manufacturing scenario. The algorithm design fits intuitively into the CPS-IIoT paradigm without exacting any additional infrastructure, and is a demonstration of how the paradigm can be effectively deployed. Due to the decentralized nature of the algorithm, its runtime scales with complexity of the problem in terms of number of machines; and the runtime for complex cases is of only a few minutes. The supremacy of the devised algorithm is demonstrated over conventional centralized heuristics such as Memetic Algorithm and Particle Swarm Optimization. © 2017 Elsevier Ltd