[STAT/AP] Melvin Drent: Real-Time Integrated Learning and Decision Making for Cumulative Shock Degradation

20 November 2023 15:45 till 16:45 - Location: Timmanzaal LB.01.170 | Add to my calendar

Unexpected failures of high-tech equipment such as medical equipment and lithography systems can have severe consequences and costs. Such unexpected failures can be prevented by performing preventive replacement based on real-time degradation data. We study a component that degrades according to a compound Poisson process and fails when the degradation exceeds the failure threshold. An online sensor measures the degradation in real time, but interventions are only possible during planned downtime.  The degradation parameters vary from one component to the next but cannot be observed directly; the component population is thus heterogeneous. These parameters must therefore be learned by observing the real-time degradation signal. We model this situation as a Bayesian Markov decision process (MDP) so that decision making and learning are integrated. We collapse the information state space of this MDP to three dimensions so that optimal policies can be analyzed and computed tractably. We characterize the optimal replacement policy as a state dependent control limit, where this control limit increases with age but may decrease as a result of other information in the degradation signal. Numerical case study analyses reveal that integration of learning and decision making leads to cost reductions of 10.50% relative to approaches that do not learn from the real-time signal and 4.28% relative to approaches that separate learning and decision making. We also demonstrate the effectiveness in practice with a case study on interventional x-ray machines. This is a joint work with Collin Drent, Joachim Arts, and Stella Kapodistria.