Mixed Integer Linear Programming Model for Scheduling Burn-in Ovens to Minimize Total Weighted Earliness/Lateness

Abstract

This study deals with scheduling of Burn-in Oven (BO), a Batch Processing Machine (BPM) in the testing stage of Semiconductor Manufacturing (SM) process. We consider a new problem configuration, close to real-life machining environment, wherein multiple BOs having different capacity are employed parallelly to process the variety of jobs. Jobs have non-agreeable release date & due date, non-identical processing time and non-identical job size. An early completion of job at BO is penalized with earliness penalty whereas late completion of job at BO causes a lateness penalty. In the current Just in Time (JIT) manufacturing scenario, an organization aims to minimize both the penalties (cost). Hence, this study aims to minimize the total weighted earliness/lateness of the jobs. We propose a Mixed Integer Linear Programming (MILP) model for dynamic scheduling of (a) single, and (b) multiple BOs. Each of the proposed MILP are validated by developing a suitable numerical example.