Improving and Maintaining the Operational Efficiency of a Semiconductor Equipment

The present work addresses an operational inefficiency problem at a semiconductor equipment manufacturing warehouse, Varian Semiconductors Associates and Equipment (VSEA). This problem is important because if unresolved, the warehouse is unable to meet the part delivery time target of 24 hours during the busy period. The downstream effects of the late part delivery are delayed production schedules and in the worst case scenario, a missed shipment to the customer, which is very costly. In order to improve the efficiency of the warehouse so as to consistently deliver parts on time, the picking efficiency needs to be enhanced. Parts are primarily picked from two types of storage locations GL shelves and Vertical Lift Modules (VLMs). The picking efficiency can be improved by the simultaneous reduction in workload on GL and improvement in the VLM picking efficiency. The first part of this thesis focuses on improving the picking inefficiency at the VLMs by employing a more efficient picking method. From our time study, we find that the pick-and-consolidate (parallel picking) is more efficient than pick-and-pass (sequential picking). The average makespan time savings per order by pick-and-consolidate is 8% (20 minutes). The second part of this paper discusses what is required to maintain a high VLM picking efficiency. New metrics to measure the workload distribution of the VLMs and the average flow time per order are proposed. Three dynamic slotting methods that maintain a balanced workload distribution across the VLMs without the need for periodic review are also examined. The methods are evaluated based on how balanced is the workload distribution across the VLMs and the cost of implementation. Thesis Supervisor: Stephen C. Graves Title: Abraham J. Siegel Professor of Management Science