Clustered Multi-Task Sequence-to-Sequence Learning for Autonomous Vehicle Repositioning

Clustered multi-task learning, which aims to leverage the generalization performance over clustered tasks, has shown an outstanding in various machine learning applications. In this paper, a sequence-to-sequence (CMSL) for autonomous vehicle systems (AVSs) large-scale semiconductor fabrications (fab) is proposed, where AVSs are widely used wafer transfers. Recently, as fabs become larger, repositioning of idle vehicles they may be requested significant challenge because inefficient balancing leads transfer delays, resulting production idleness. However, existing mainly controlled by human operators, and it difficult such guarantee efficiency. Further, we should handle small data problem, insufficient irregular time-varying manufacturing environments. The main purpose study examine CMSL-based predictive control maximize utilization. We conducted experimental evaluation compare prediction accuracy CMSL with methods. case real largescale plant, demonstrated that proposed approach outperforms approaches terms efficiency