Schedule Optimization for Construction in Manufacturing Environments

Construction in a semi-conductor manufacturing facility involves the constant construction, demolition and modification of thousands of machines that enable raw silicon to be transformed to a wafer with over 1.4 billion transistors. While maintenance is always required to operate such large facilities, a majority of the construction follows a cyclical pattern of a two year cycle. The ability to handle the construction loads of over 3,000 machines in a 6-12 month period requires a specialized construction workforce that is able to meet the strict quality requirements of working in a class 1 cleanroom environment. Within a given construction ramp, the trade headcount can rise from hundreds to low thousands during the peak periods. Baseline schedules are usually created around the technology demands without added cost or inefficiencies to the construction contractors. These inefficiencies range from slipped schedules to added cost due to overtime of rework requirements. The model we propose treats each construction activity as an individual project and aims to minimize the total cost of labor during the technology ramp while also minimizing the amount of labor resources that are hired and fired. Labor resources represent the largest risk to the program from a cost, schedule and quality focus