Setup Change Scheduling for Semiconductor Packaging Facilities Using a Genetic Algorithm With an Operator Recommender

Semiconductor manufacturers are increasingly assembling multiple chips into a single package to maximize the capacity of flash memories. Multiple-chip products (MCPs) require repetitive visits to assembly stages and incur frequent setup changes. As utilization of packaging facilities decreases due to the introduction of MCPs, research on scheduling of packaging facilities is becoming more important than ever. In this paper, we propose a novel framework to find a good schedule for semiconductor packaging facilities by focusing on bottleneck stages while satisfying practical operational constraints. A genetic algorithm-based sequence optimizer is employed, and construction and performance evaluation of a schedule are separately addressed by a simulator. Furthermore, a recommender is proposed to accelerate convergence of the optimizer. Experimental results show that the proposed approach performs better than the other existing methods while successfully reducing computation time.