Operationalizing Industrial Sensor Data for Scalable Asset Management in Energy Systems: Advancing Methods at the Intersection of Prediction and Decision Making
Murat YILDIRIM
Sensor-driven maintenance and operations scheduling in energy systems revolves around coordinating fleet-level electricity production with sensing and asset monitoring to help support maintenance decisions and control asset loading. What makes this setting interesting is the presence of unique interactions and dependencies among energy assets, which are typically driven by different physical phenomena and complex constraints such as power flow, degradation, and operational limits. In this talk, I will present a unified framework that embeds predictive degradation models pertaining to the generation assets within decision optimization models to jointly solve operations and maintenance in a variety of energy system settings. I will demonstrate mathematical programming and attention-based mechanisms to address challenges associated with uncertainty modeling, scalability, and privacy. Using classic benchmarks from the IEEE community coupled with real-world sensor data, I will illustrate some of the considerable cost and reliability improvements relative to existing state-of-the-art approaches.
Short Bio
Murat Yıldırım is an Associate Professor in the Department of Industrial and Systems Engineering at Wayne State University. Prior to joining Wayne State, he worked as a postdoctoral fellow at the Georgia Institute of Technology (2016-2018). He obtained a PhD degree in Industrial Engineering, MS degree in Operations Research, and BS degrees in Electrical Engineering and Industrial Engineering from Georgia Institute of Technology. Dr. Yildirim's research interest lies in advancing the integration of mathematical programming and data analytics in various application domains. Specifically, he focuses on the modeling and the computational challenges arising from the integration of real-time inferences generated by advanced data analytics and simulation into largescale decision optimization models used for optimizing and controlling networked systems. To date, Dr. Yildirim’s research has been supported through funding from NSF, DoE, Michigan Translational Research and Commercialization, and Ford Motor Company.
Venue
Friday, November 14th, 2025, 5:00 pm
Online
The link for the seminar is: https://teams.microsoft.com/meet/35755257171668?p=BUuJ9JTcZCufuFWrFz
Meeting ID: 357 552 571 716 68
Passcode: Mo7eJ3fx