Variable neighbourhood search for job scheduling with position-dependent deteriorating processing times

Abstract

This article studies the problem of job scheduling in hand-intensive manufacturing with processing times that deteriorate depending on the sequence in order to minimize the maximum delay of the set of jobs. In the literature, only a small number of academic works has considered the assumption of non-stationary workers during the planning horizon. This problem is NP-hard since it is as an extension of the parallel machine scheduling problem with makespan minimization, which is itself NP-hard. A variable neighbourhood search (VNS) algorithm is presented to solve this problem; its efficiency is evaluated through an extended set of numerical experiments with random-generated datasets. Results are compared with an exact method based on mathematical programming. The results show that the VNS algorithm provides good solutions in reasonable computing times even for large problems.