PhD defence S. (Sha) Zhu

My research focuses on the spare parts demand forecasting and inventory management problem. First, we conduct research on intermittent demand forecasting. We propose an improved spare parts demand forecasting method using extreme value theory (EVT). Our results show that the improved method is competitive with the state-of-the-art methods for a range of demand distributions. We next focus on the advance demand information provided by the on-condition maintenance. We study the value of such advance demand information in the decision of spare parts inventory management. We also study the spare parts ordering problem against this background of shutdown maintenance project planning. Differentiate from standard spare parts inventory control theory in which one assumes that decisions for each part separately can be made and that either a fill-rate target is set or a stockout penalty is given, in our setting there is no clear out-of-stock costs or fill-rate target for the spare part and the consequence of one part being out of stock may also depend on the availability of other parts. Unlike unimodal activity durations which are widely considered in literature, we consider bimodal (with two peaks) activity durations which make the total project time multimodal rather than unimodal. We find that scenario based heuristics give an acceptable approximation, while the standard project management approach based on the widely used critical path method gives poor solution in our setting. We employ techniques and methods from both the operations research and statistics fields to aid in these issues.

The PhD defences do not take place publicly in the usual way in the Senate Hall at campus Woudestein or in the Professor Andries Querido Room at the Erasmus MC. The candidates will defend their dissertation either in a small group or online.