Thom Badings
I am a postdoctoral research associate with the Oxford Control and Verification Group at the University of Oxford, working with Prof. Alessandro Abate. Previously, I was a PhD student in the Department of Software Science at Radboud University in Nijmegen, the Netherlands, under supervision of Prof. Nils Jansen and Prof. Marielle Stoelinga. I defended my PhD thesis titled “Robust Verification of Stochastic Systems: Guarantees in the Presence of Uncertainty” in March 2025 (Cum Laude, awarded to around 5% of PhD theses in the Netherlands).
I am a research associate in the Erlangen AI Hub on the mathematical foundations of AI. During my PhD, I was a member of PrimaVera, an academic consortium on the topic of predictive maintenance.
You can reach me at thom.badings?cs.ox.ac.uk.
Research Goal
There is a quest for reliable autonomy in control systems: An autonomous drone must not crash under adverse weather, and a smart energy grid must remain stable despite fluctuating demand and supply. Growing complexity and uncertainty about systems and their environment have created a gap between the systems we build and the systems we can reliably control. My long-term research goal is to close this gap.
Achieving this goal requires reasoning about control decisions in systems. This reasoning uses, in particular, dynamical systems and Markov decision processes. These models are studied across control theory, formal methods, and artificial intelligence. My research spans and has impact in all three communities.
news
2025
October
- Excited to be an invited speaker at the GT-Vérif 2025 annual meeting (Nov 5-7), hosted at Université Paris-Est - Créteil in Paris. Very much looking forward to share our latest work on control design of uncertain and stochastic systems with formal guarantees.
July
- Two papers accepted at CDC 2025! Both on advancing the theoretical foundations of abstraction-based policy synthesis for stochastic systems: one approach data-driven and the other model-based.
- "Data-Driven Abstraction and Synthesis for Stochastic Systems with Unknown Dynamics" with Mahdi Nazeri, Anne-Kathrin Schmuck, Sadegh Soudjani, and Alessandro Abate
- "Probabilistic Alternating Simulations for Policy Synthesis in Uncertain Stochastic Dynamical Systems" with Alessandro Abate
April
- Our paper “Learning-Based Verification of Stochastic Dynamical Systems with Neural Network Policies”, which is joint work with Wietze Koops, Sebastian Junges, and Nils Jansen, has been accepted for presentation at CAV 2025!
March
- I am extremely happy to share that, on March 27th, I was awarded my PhD cum laude! I want to thank my advisors, Nils and Mariëlle, the entire manuscript committee and all other opponents, as well as everyone else present at the defense, for making this a truly memorable day.
- Our paper “Data-Driven Yet Formal Policy Synthesis for Stochastic Nonlinear Dynamical Systems”, coauthored by Mahdi Nazeri, Sadegh Soudjani, and Alessandro Abate has been accepted for presentation at TACAS 2024. Especially congrats to Mahdi for the first paper part of his PhD!
February
- I’m proud to share that our paper “Integrating Expert and Physics Knowledge for Modeling Heat Load in District Heating Systems”, written together with Francisco Souza, Geert Postma, and Jeroen Jansen has been accepted for publications in the IEEE Transactions on Industrial Informatics!
2024
November
- I am excited to share that I have started a new position as a postdoctoral research associate in the Oxford Control and Verification Group (OXCAV) at the University of Oxford!
April
- This June, we will present our paper “A Stability-Based Abstraction Framework for Reach-Avoid Control of Stochastic Dynamical Systems with Unknown Noise Distributions” at the European Control Conference (ECC) in Stockholm.
2023
November
- Our paper “CTMCs with imprecisely timed observations” has been accepted for presentation at TACAS 2024!
- In the past two weeks, we have presented our work on robust abstraction-based control under uncertainty at the peer-reviewed workshops BNAIC and FMAS. Really nice venues, with lots of interesting talks!
September
- In around two weeks, we will present our paper “Formal Controller Synthesis for Markov Jump Linear Systems with Uncertain Dynamics” at QEST 2023 (September 20-22). Great collaboration with Luke Rickard, Licio Romao, and Alessandro Abate from the University of Oxford!`
April
- Good news, we will present our recent paper “Efficient Sensitivity Analysis for Parametric Robust Markov Chains” this summer at CAV 2023!
January
- Our paper “Robust Control for Dynamical Systems with Non-Gaussian Noise via Formal Abstractions” has been published in the Journal of Artificial Intelligence Research (JAIR), in their ‘award-winning paper track.’
2022
November
- Our paper “Probabilities Are Not Enough: Formal Controller Synthesis for Stochastic Dynamical Models with Epistemic Uncertainty” has been accepted for presentation at AAAI 2023.