Bottleneck-centric pull and push allocation and sequencing of wet-bench and furnace tools

Furnace tool group is often a bottleneck of semiconductor fabrication, which makes effective scheduling of furnace tools and its upstream tool group, wet-bench, important to operation efficiency. Major challenges come from not only the problem complexity such as batching requirements, internal tool sequence, and waiting time limitations but also a reasonable computation time for rescheduling. This paper develops a novel solution scheme, a push-and-then-pull bottleneck centric integrated allocation and sequencing method (P&P BCIASM), to integrate the sequencing and allocation for wafers processing by wet bench and furnace tools effectively. Furnace tool allocations, first derived from a coarse wet-fine furnace model, are viewed as pull demands for allocating and sequencing wet-bench tools. The allocations of wet tools thus obtained are considered as push supplies for refining the scheduling of furnace tools. A commonly available optimization tool is skillfully applied to solution finding. Test results over problem instances extracted from real fab data demonstrate that P&P BCIASM leads to near-optimal schedules, 2.56% higher utilization in average than a hybrid allocation scheme of optimization and manual adjustment, and 15.04% mean waiting time reductions. The violation of waiting time limitation is reduced to none. The computation time in average is within one minute, indicating a strong potential for applications to both large size problems and dynamic allocations.