Capacitated disassembly scheduling with parts commonality and start-up cost and its industrial application

This paper discusses the capacitated disassembly scheduling with parts commonality and start-up cost, which has broad applications in remanufacturing and many other production systems. The problem is formulated as a mixed-integer programming model. A Lagrangian heuristic is developed, where the lower bound problem reduces to a linear model, and the upper bound can be obtained with a two-stage procedure in polynomial time. The algorithm is tested on randomly generated problems. The results reveal that the approach outperforms CPLEX significantly for all the test instances, and for real-scale problem it provides sufficiently good solutions within acceptable time. Sensitivity studies on capacity, parts commonality and start-up cost are also conducted, which provide some useful insights for industrial managers. Moreover, a real-world case from a valve disassembly factory is studied, which illustrates the application of the proposed approach and gives managerial insights. Finally, conclusions are drawn and some potential extensions are discussed.