(A) scheduling algorithm for diffusion process with limited waiting time constraints in a semiconductor wafer fabrication facility

Abstract

This study focuses on the problem of scheduling wafer lots with limited waiting times between pairs of consecutive batch operations of semiconductor fabrication facility. For example, after a cleaning operation for a wafer lot is completed, the next operation should be started within a certain time period on the diffusion workstation so that cleaned surfaces of the wafers do not get contaminated or natural oxidation does not occur. Existed various heuristic algorithms with limited waiting time allow some time violations but those algorithms overlook the special cases of which the time constraint is uncompromising. Therefore in this paper, we develop a new effective heuristic algorithm which is satisfied with waiting time constraints perfectly. Our objective is the minimizing total tardiness of orders and we develop a heuristic algorithm by using CSP (Constraint Satisfaction Problem) algorithm for batching and scheduling wafers considering the states of the workstations for the operation and its successor or predecessor operations. To evaluate performance of the suggested scheduling method, we perform computational experiment using real factory data as well as randomly generated data sets. Results of the experiments show that the suggested method performs better than those of other research and the one that has been used in practice.