Thesis defense of Qianru Ge, June 25, 2018


On Monday, June 25, 2018, Qianru Ge will defend her PhD thesis “Reliability optimization and system redesign under uncertain failure rates”.
This thesis has been supervised by prof.dr.ir. G.J.J.A.N. van Houtum, prof.dr.ir. I.J.B.F. Adan and dr. Z. Atan.
The ceremony will take place in the Collegezaal 4 of the Auditorium at the Eindhoven University of Technology, at 16:00 hrs.

Abstract
The serious consequences of the downtime of capital goods and the difficulty of maintaining these goods put a lot of pressure on the Original Equipment Manufacturers (OEMs) who are not only responsible for manufacturing capital goods but also for taking care of their maintenance to guarantee high availability under service contracts. To assist OEMs with their challenge, we develop decision support models to find the optimal reliability/redesign policies which minimize the life cycle cost of the capital goods. We study performance-based contracts with uncertain failure rates.

In our first model, we determine the optimal design of critical components in a serial system. All components are subject to failures with uncertain rates. We develop an approximate evaluation method to determine the best design for every critical component.

We also study the reliability optimization problem for a group of identical systems sold under a service contract that is divided into several subperiods with a threshold downtime per each subperiod. We propose an exact method to find the optimal reliability levels of the components.

The first two models optimize the design decisions. In addition, for a given initial design, we provide two modes to help the OEMs with their redesign decisions. In the first redesign model, we consider a system that is subject to random failures with an uncertain failure rate. Our model selects the optimal redesign investment level for each period. In our second redesign model, we consider multiple failure modes and decide how to distribute the redesign investment among different failure modes.