An artificial immune system algorithm for optimizing the distributed production scheduling in the semiconductor assembly industry

The semiconductor industry faces one of the most challenging and rapidly evolving problem in each passing time. The importance of satisfying customer demands and achieving higher profits have increased the attention towards production scheduling in semiconductor assembly areas. Combination of productivity and flexibility provided by semiconductor assembly areas have attracted many research efforts for several years. Additionally, recent globalization events have encouraged decentralized efforts in the semiconductor assembly industry by implementing distributed systems. With respect to the previously mentioned problems, several approaches have been introduced namely Petri nets, Ant Colony, Genetic Algorithm, Intelligent Agents, and hybrids of the latter. A new model, using artificial immune system (AIS) algorithm, slightly modified to conform to the manufacturing constraints, has been proposed and tested with an industrial dataset to establish its effectiveness. AIS algorithm is chosen due to its inherent self-learning capability and memory incubation features. The results from the experiments conducted show that the proposed algorithm is effective for the aforementioned problems where the production schedule efficiency increased from 30% to 60%.