Hong, Cheolhui
(Korea Advanced Institute of Science and Technology (KAIST),Department of Industrial and Systems Engineering,Daehak-ro 291,Daejeon,South Korea)
,
Lee, Tae-Eog
(Korea Advanced Institute of Science and Technology (KAIST),Department of Industrial and Systems Engineering,Daehak-ro 291,Daejeon,South Korea)
Due to extreme quality control requirements, recent cluster tools for semiconductor manufacturing processes tend to have an equipment front-end module (EFEM), where wafer cassettes are loaded and unloaded, and aligned before processing at chambers and cooled before returned to the loadport. Strict w...
Due to extreme quality control requirements, recent cluster tools for semiconductor manufacturing processes tend to have an equipment front-end module (EFEM), where wafer cassettes are loaded and unloaded, and aligned before processing at chambers and cooled before returned to the loadport. Strict wafer treatment requirements in an EFEM make tool operation complicated, and often cause logical problems like deadlocks. We therefore propose a logical control method for dealing with such complexity. We first present a number of extended finite state machines (EFSM) models. They specify proper tool operation at the wafer aligner, cooler, and atmosphere robot of the EFEM as well as the wafer processing chambers and the vacuum handling robot. Each EFSM includes a control specification that enforces the sequence of wafer processes at chambers, aligning or cooling recipes at the EFEM. To prevent a deadlock due to complicated operation and inefficient operation, by following a part of the supervisory control theory, we make a synchronous product of the extended state machines to form an integrated global state machine and trim out deadlock-inducing states. We use the resulting state machine as a supervisory controller that ensures proper tool operation. To confirm proper tool control, we then experiment simple scheduling strategies prior to robust scheduling for further performance optimization by tool simulation based on the EFSM models and their supervisory controller.
Due to extreme quality control requirements, recent cluster tools for semiconductor manufacturing processes tend to have an equipment front-end module (EFEM), where wafer cassettes are loaded and unloaded, and aligned before processing at chambers and cooled before returned to the loadport. Strict wafer treatment requirements in an EFEM make tool operation complicated, and often cause logical problems like deadlocks. We therefore propose a logical control method for dealing with such complexity. We first present a number of extended finite state machines (EFSM) models. They specify proper tool operation at the wafer aligner, cooler, and atmosphere robot of the EFEM as well as the wafer processing chambers and the vacuum handling robot. Each EFSM includes a control specification that enforces the sequence of wafer processes at chambers, aligning or cooling recipes at the EFEM. To prevent a deadlock due to complicated operation and inefficient operation, by following a part of the supervisory control theory, we make a synchronous product of the extended state machines to form an integrated global state machine and trim out deadlock-inducing states. We use the resulting state machine as a supervisory controller that ensures proper tool operation. To confirm proper tool control, we then experiment simple scheduling strategies prior to robust scheduling for further performance optimization by tool simulation based on the EFSM models and their supervisory controller.
※ AI-Helper는 부적절한 답변을 할 수 있습니다.