Bi-objective optimization of a job shop with two types of failures for the operating machines that use automated guided vehicles

Highlights: • A bi-objective reliability optimization via simulation is proposed for a job shop. • The failure times of the parallel machines in a shop follow either an exponential or a Weibull distribution. • A simulation approach is taken to estimate the reliability of the shops having machines with Weibull failures. • NSCS and MOTLBO algorithms are designed to solve the problem. • AHP-TOPSIS is used to rank the algorithms in terms of five performance metrics. Reliability of machinery and equipment in flexible manufacturing systems are among the most important issues to reduce production costs and to increase efficiency. This paper investigates the reliability of machinery in job shop production systems, where materials, parts, and other production needs are handled by automated guided vehicles (AGV). The failures time of the parallel machines in a given shop follow either an exponential or a Weibull distribution. As there is no closed-form equation to calculate the reliability of the shop in the Weibull case, a simulation approach is taken in this paper to estimate the reliability. Then, a bi-objective nonlinear optimization model is developed for the problem under investigation to maximize shop reliability as well as to minimize production time, simultaneously. In order to assess the efficacy of the proposed model, some random instances are generated, based on which two meta-heuristic algorithms called non-dominated sorting cuckoo search (NSCS) and multi-objective teaching–learning-based optimization (MOTLBO) are designed. Finally, to evaluate and compare the effectiveness of the proposed solution algorithms, an efficient solution AHP-TOPSIS technique is utilized.