A fast rescheduling method in semiconductor manufacturing allowing for tardiness and scheduling stability

Abstract

In semiconductor manufacturing, rescheduling problems are extremely difficult to solve in real time due to the high frequency of disturbances that occur approximately every minute. This paper proposes a new approach to online manufacturing rescheduling. Unlike the traditional methods to have a scheduling process again, we (1) revise the existing schedule to keep high schedule stability based on message passing rescheduling with operation sorting; and (2) during the manufacturing process, improve the performance of the revised schedule with the introduction of a quick local search on semi-critical paths. In actual problems with about 200,000 processes, this method can effectively accommodate disturbances in less than 1 second, and a better schedule can be obtained in less than 1 minute. This method has been demonstrated to be more effective than conventional dispatching-rule methods, some of which have been actually applied in many facilities, because it offers higher schedule stability and fewer violations of due dates.

Artificial Neural Networks in Production Scheduling and Yield Prediction of Semiconductor Wafer Fabrication System

“…Therefore, production scheduling is required to optimize the operation of SWFS and has been reviewed by Uzsoy and his colleagues [3]. SWFS operates in uncertain dynamic environments, facing with a lot of disturbances, such as machine failure, a lot of rework and rush orders [4]. Production rescheduling has been viewed as an efficient approach in responding to these uncertainties raised by the external environment and internal conditions of production [5].…”

Single Machine Oriented Match-Up Rescheduling Method for Semiconductor Manufacturing System

“…In this case, rescheduling is necessary to react to disturbances. That is to say, manufacturers should not only create the best schedules but also react quickly to disturbances and revise original schedules in small computational cost [3]. According to static scheduling and dynamic scheduling categories of scheduling philosophy, rescheduling belongs to the dynamic scheduling, which pays more attention to the needs of practical environments, and attempts to make a schedule more applicable.…”

“…Based on the literature review [3,[8][9], the predictive-reactive rescheduling strategies can take account of both optimization and flexibility by a scheduling/rescheduling procedure. A partial repair rescheduling method is superior to generation rescheduling in its potential to maintain schedule stability with reasonable computational cost.…”