Optimising the maintenance strategy for a multi-AGV system using genetic algorithms

Automated Guided Vehicles (AGVs) are playing increasingly vital roles in a variety of applications in modern society, such as intelligent transportation in warehouses and material distribution in automated production lines. They improve production efficiency, save labour cost, and bring significant economic benefit to end users. However, to utilise these potential benefits is highly dependent on the reliability and availability of the AGVs. In other words, an effective maintenance strategy is critical in the application of AGVs. The research activity reported in this paper is to realise an effective maintenance strategy for a multi-AGV system by the approach of Genetic Algorithms (GA). To facilitate the research, an automated material distribution system consisting of three AGVs is considered in this paper for methodology development. The movement of every AGV in the multi-AGV system, and the corrective and periodic preventive maintenances of failed AGVs are modelled using the approach of Coloured Petri Nets (CPNs). Then, a GA is adopted for optimising the maintenance and associated design and operation of the multi-AGV system. From this research, it is disclosed that both the location selection of the maintenance site and the maintenance strategies that are adopted for AGV maintenance have significant influences on the efficiency, cost, and productivity of a multi-AGV system