Predictive aircraft maintenance: From predicting the failure time to optimizing the maintenance schedule

Please let one of us (Ruben, Olga, or Ece) know if you would like to have a meeting with the speaker after the seminar or if you would like to attend the seminar online. 

The current aircraft maintenance practice mostly consists of preventive maintenance: Most aircraft components either undergo frequent inspections or are maintained after a fixed period of time. These frequent inspections and time-based replacements ensure the safety of the aircraft, but also make aircraft maintenance expensive.

My research therefore focuses on a new maintenance strategy, called predictive maintenance. In aircraft, many sensors are installed around the aircraft components. In predictive maintenance, the measurements of these sensors are used to predict the time left until the failure of each individual component. This is called the Remaining Useful Life (RUL).  Subsequently, these RUL predictions are integrated in the aircraft maintenance planning. Here, the challenge is to take the uncertainty of the RUL predictions into account. The ultimate aim of predictive maintenance is to only maintain an aircraft component just before it fails, thus making aircraft maintenance very efficient.

In this seminar, I will illustrate the first step (predicting the failure time of aircraft components) and the second step (optimizing the maintenance schedule with these uncertain predictions) with two different case studies.

About Ingeborg de Pater

Ingeborg de Pater recently finished her PhD in predictive aircraft maintenance at the TU Delft, and this month, she will start as an assistant professor at Delft on the same topic. Before this, she studied econometrics with a master's in OR at Erasmus University, so she is looking forward to being back again for a day!