Research
Developing data-driven and probabilistic frameworks for risk-informed inspection, monitoring, and maintenance of infrastructure systems under uncertainty.
Focus Areas
01
Markov decision processes and partially observable MDPs for adaptive inspection, monitoring, and maintenance planning under uncertainty. Safety calibration and risk-informed optimization of deteriorating engineering systems.
02
Multi-agent and hierarchical RL frameworks for scalable, coordinated management of large infrastructure networks. Emphasis on interpretability and policy robustness in safety-critical settings.
03
Bayesian neural networks and dynamic Bayesian networks for probabilistic inference from engineering data. Active learning methods for efficient experiment design and surrogate-model-based reliability analysis.
04
Risk-based inspection planning, fatigue and fracture reliability, and safety calibration for deteriorating structural components under realistic loading and environmental conditions.
05
Long-horizon, risk-aware optimization of networked and deteriorating systems, spanning offshore wind farms, critical civil infrastructure, and urban building stock.
06
Structural health monitoring, virtual load monitoring, and integrity management for offshore wind substructures. Application of probabilistic methods to fatigue assessment under operational and environmental uncertainty.
Invited & Conference Talks
Optimization under Uncertainty — MS139
WCCM-ECCOMAS 2026 · Munich, Germany · Jul 2026 (Co-organizer)
IMP-MARL: A Suite of Environments for Large-Scale Infrastructure Management via Multi-Agent Reinforcement Learning
NeurIPS 2023 · New Orleans, USA · Dec 2023
Interpretation and Analysis of Deep Reinforcement Learning Driven Inspection and Maintenance Policies
ICASP14 · Dublin, Ireland · Jul 2023
Managing Offshore Wind Turbines through Markov Decision Processes and Dynamic Bayesian Networks
ICOSSAR 2022 · Shanghai, China · Jun 2022
POMDP Based Maintenance Optimization of Offshore Wind Substructures including Monitoring
ICASP13 · Seoul, South Korea · May 2019
Add further talks by copying a <li class="talk-item"> block above.
Selected Projects
PhairywinD
2020–2025Belgian Energy Transition Fund (FPS Economy)
Probabilistic framework for fatigue reliability, inspection planning, and life-cycle management of offshore wind support structures under realistic loading and environmental conditions.
MAXWind
2020–2025Belgian Energy Transition Fund (FPS Economy)
Maximizing the operational lifetime of offshore wind turbines through advanced structural health monitoring, virtual load sensing, and data-driven integrity assessment methods.
DE-CIST
2023–2025Municipality of Rotterdam · Google
City-scale building energy retrofit planning tool combining Bayesian neural networks and uncertainty quantification to prioritize interventions across Rotterdam's building stock.
IMP-MARL
2023Open-source · DTU / ULiège
Open-source benchmark suite of cooperative multi-agent RL environments for large-scale infrastructure management planning, motivated by offshore wind inspection and maintenance problems.
Research Trajectory