Optimal specialization of a maintenance workforce

Abstract

This article develops an analytical method for determining an optimal specialization strategy for a maintenance workforce. The method assumes that maintenance tasks are generated by a system of statistically identical machines that experience random malfunctions and require periodic service. The impact of alternative workforce structures on system performance is evaluated with a queueing network model. Markov decision analysis is employed to determine an optimal assignment of maintenance personnel to pending tasks as the network status varies over time. A linear programming algorithm is derived to enable simultaneous optimization of specific assignment decisions and the overall workforce structure. A manufacturing example demonstrates the applicability of the method to many industrial contexts. The method is also applied to the problem of maximizing military aircraft sortie generation subject to a constraint on maintenance personnel expenditure.