Scheduling single-blade cluster tools with time window constraints

In semiconductor manufacturing, finding an efficient way for scheduling a cluster tools is crucial for productivity improvement and cost reduction. In this paper, we investigate the cycle scheduling problem with time window constrains in cluster tools. The objective is to minimize the cycle time. To solve the problem, we present a searching algorithm which based on linear programming model and the collision detection strategy. The algorithm uses the fundamental period as controlling parameters to generate all the possible permutations of robot action sequence in the cycle. A linear programming model is then used to find the optimal schedule for each of these permutations based on our constraints. To avoid the collision of robot action, we solve it with slide and stretch strategy. The experiments of Spin Coater demonstrate the effectiveness and efficiency of the proposed algorithm.