(An) advanced dynamic route guidance algorithm of overhead hoist transport systems in semiconductor fabs

Abstract

This paper presents an advanced route guidance algorithm of Overhead Hoist Transport (OHT) in semiconductor fabs. The state-of-art of dynamic route guidance algorithm is by Hwang and Jang (2018), which have shown outperforming results compared to previous algorithms. However, it does not function properly under the situation with abrupt transportation pattern. In this method, traffic conditions are reflected only after the OHT has finished the travel, making it difficult to find efficient routes when a large number of requests are concentrated at once, such as batch processing. we propose an Active update method to predict and reflect the traffic information of OHTs in advance, and introduces the methodology of Reward clipping, Active adjustment to enable more efficient decision making. The improved performance was verified with Applied Materials AutoMod™ (version 14.0) simulation software.