Capacitated disassembly scheduling under stochastic yield and demand

The disassembly process in remanufacturing has attracted increasing attention in recent years due to the high importance of environmental issues. The paper studies a capacitated single-item multi-period disassembly scheduling problem with random yields and demands in which procured, returned items (root items) are disassembled into components or parts (leaf items) to satisfy their demands in each period. The problem is formulated as a mixed integer nonlinear program (MINLP). Notably, a chance constraint is introduced to ensure that the probability of satisfying the demand is greater than a predetermined service level, and then is approximated by a second-order cone constraint. An outer approximation (OA) based solution algorithm is proposed to solve the resulting model. Furthermore, a special case that has uniformly-distributed yield and normally-distributed demand is considered given a closed-form formulation. Extensive numerical experiments demonstrate that the proposed algorithm can achieve converged optimal solutions within much less CPU time compared with the well-known solver BONMIN. Furthermore, a scheduling process of a valve producer is also conducted to demonstrate the application in practice. In the end, managerial insights are explored and future research directions are outlined.