Optimal Scheduling by Workload Analysis for Cluster Tools

Abstract

Cluster tools, each of which consists of several single-wafer processing chambers, a wafer handling robot and loadlocks for wafer cassette loading, have been widely used for various wafer fabrication processes. Recently, they tend to be connected together or more chambers are serially configured in a tool with multiple robots in order to reduce lot transfers between tools and increase the tool throughput by parallelizing a long process step with multiple chambers. One of the objectives of this research is to provide a unified cyclic scheduling method for such sequentially connected tools. Another objective is to discuss optimal scheduling of transient cycles for cluster tools. Scheduling problems of transient cycles such as start-up cycles and close-down cycles have not been well addressed compared to those of steady full cycles in which cluster tools can repeat identical work cycles. However, the proportion of the transient cycles has become larger than before due to smaller lot size. For these problems, we use the concept of the workload, the minimal time to produce a wafer at the resource, and extend the workload to the problems for sequentially connected tools and transient cycles.

In Chapter 2, we examine cyclic scheduling problems of sequentially connected cluster tools with separated input and output modules, which include linear cluster tools and in-line multiple cluster tools. We first compute workloads for each process step and each robot, of which the maximum provide a lower bound of the tool cycle time. We then identify an assignment rule of assigning the chambers to the process steps that makes the tool cycle time independent of the order of using the parallel chambers. We also propose robot task sequences for each component cluster tool, which are modified from the well-known backward sequence or swap sequence. We prove that the modified backward sequence and the swap sequence are optimal for single-armed component tools and dual-armed component...