Production scheduling/rescheduling in flexible manufacturing

TLDR

Scheduling in flexible manufacturing systems has been extensively studied, yet uncertainties and modeling limits cause deviations that make reactive rescheduling essential. The study aims to address four common uncertainties—machine breakdowns, priority changes, rush orders, and cancellations—by incorporating real‑time shop status into rescheduling. Genetic algorithms generate an initial schedule, and the proposed method selectively revises only affected operations to integrate uncertainty handling with existing scheduling approaches. The resulting algorithms efficiently update only necessary operations, enhancing overall system efficiency when combined with existing scheduling methods.

Abstract

Scheduling of production in flexible manufacturing systems (FMSs) has been extensively researched over the past years and it continues to attract the interests of both academic researchers and practitioners. The generation of new and modified production schedules is becoming a necessity in today's complex manufacturing environment. Genetic algorithms are used in this paper to obtain an initial schedule. Uncertainties in the production environment and modelling limitations inevitably result in deviations from the generated schedules. This makes rescheduling or reactive scheduling essential. Four different types of uncertainties that normally cause discrepancies between the actual output and the planned output are considered in this paper. These include unforeseen machine breakdowns, increased order priority, rush orders arrival and order cancellations. In this paper, the current status of the shop is considered while rescheduling. The proposed algorithms revise only those operations that must be rescheduled and can, therefore, be used in conjunction with the existing scheduling methods to improve the efficiency of flexible manufacturing systems.