Juseong Lee is an assistant professor at Operations Planning Accounting & Control (OPAC) group, the Department of Industrial Engineering & Innovation Sciences, Eindhoven University of Technology (TU Eindhoven). He holds a PhD in Aerospace Engineering from Delft University of Technology (TU Delft) and an MSc in Aerospace Engineering from the Korea Advanced Institute of Science and Technology (KAIST).
JJuseong’s research focuses on integrating data-driven models and AI for decision-making in maintenance and manufacturing. He works to realize the predictive maintenance of industrial assets using data-driven condition monitoring and physics-informed AI models for diagnostics and prognostics. He also works to realize smart manufacturing, such as data-driven quality management and automated product design optimization. Currently, he is working on data-driven diagnostics/prognostics, digital twins for data generation, and explainable decision support tools. Check his recent works.
Before joining TU/e, he worked on the optimal design of predictive aircraft maintenance using the data-driven prognostics of Remaining-Useful-Life. He participated in the European project ReMAP H2020, developing a multi-objective performance assessment framework for predictive aircraft maintenance. Check his research projects.
E-mail: J.Lee@tue.nl
Updates
- Supporting Decisions in Predictive Maintenance: Using Deep Reinforcement LearningEuropean Safety and Reliability Conference (ESREL) was full of inspiring presentations on modeling, digital twins, prognostics, optimization, and deeper analysis and practical applications on maintenance, reliability, and safety. This year, I presented how to support maintenance planners’ decisions based on probabilistic predictions of Remaining-Useful-Life. We came up with two ideas. First, Monte Carlo dropout, which prevents over-fittingContinue reading “Supporting Decisions in Predictive Maintenance: Using Deep Reinforcement Learning”
- Analysis of Emerging Challenges for Data-Driven Predictive Aircraft MaintenanceInnovation sounds cool but always involves unexpected challenges. We can safely implement new approaches only if we carefully consider new challenges in advance. Data-driven predictive aircraft maintenance is a new promising strategy to improve the efficiency and reliability of air transportation. Our recent work identifies and analyzes emerging challenges for data-driven predictive aircraft, using agent-basedContinue reading “Analysis of Emerging Challenges for Data-Driven Predictive Aircraft Maintenance”
- DRL for Predictive Maintenance using probabilistic RUL prognosticsDeep reinforcement learning (DRL) is useful for repeating decision-making based on complex input data. For predictive aircraft maintenance, it is crucial to schedule maintenance based on the Remaining-Useful-Life (RUL) predictions. Especially, when the RUL is given as a probability distribution, it is not trivial to schedule optimal maintenance. I proposed a DRL approach to decideContinue reading “DRL for Predictive Maintenance using probabilistic RUL prognostics”
- ESREL 2022, Innovation in Transportation ApplicationsI presented my work, “Designing Reliable, Data-Driven Maintenance for Aircraft Systems with Applications to the Aircraft Landing Gear Brakes,” at the 32nd European Safety and Reliability Conference (ESREL 2022). My work won the PhD Award for Innovation in Transportation Applications, which has been supported by PayPal. I am honored to receive this award from the mostContinue reading “ESREL 2022, Innovation in Transportation Applications”
- PHM Society 2022, The Best Paper Award – Second PrizeI presented my recent work “Remaining-Useful-Life Prognostics for Opportunistic Grouping of Maintenance of Landing Gear Brakes for a Fleet of Aircraft,” at the 7th European Conference of the Prognostics and Health Management Society 2022. This paper won the Best Paper Award, Second Prize. This research considers the next step after PHM. We proposed how toContinue reading “PHM Society 2022, The Best Paper Award – Second Prize”