Abstract
This study proposes a mixed integer programming approach to solve the timing and amount of capacity expansion project under uncertain demands in a high-tech industry. Such a capacity expansion project is usually composed of several sequenced phases to create new capacity and each phase involves a set of activities with precedence relationships. Given multiple demand scenarios and activity cost scenarios with known probabilities, we formulate the capacity expansion problem as a stochastic project scheduling problem, in which lead time, budget limitation and allowance of debt are taken into consideration. The goal is to decide the amount of new capacity of activities and the timing to execute the activities, with the objective of minimizing the expected costs of undersupply and oversupply to future market demands and the expected cost of executing activities. Finally, we test three different experiments by changing the probability of scenarios, the ratio of undersupply cost to oversupply cost and the availability of budget, to observe the performance and verify our model.