抄録
This paper describes a new approach for work assignment among operators using reinforcement learning. The work assignment is self-organized as the result of local interaction between machines and operators that have a reinforcement learning unit to decide their operational machines. A case study is presented for a semiconductor manufacturing in which it is difficult to find a proper work assignment plan because of large scale of the system and complex process flow. The effectiveness of the proposed method is discussed in the computer simulations results.
