Research Interest

Building safe and trustworthy cyber-physical systems

My research interests are in building cyber-physical systems, mostly AI systems, that are reliable, robust, and adaptive. More specifically, I am interested in utilizing formal methods theory and frameworks such as DSLs and probabilistic model checking to enhance safety reasoning in AI models and improve the trust between users and AI-system outputs.

Formal Methods Safety Runtime Monitoring Explainable AI Reinforcement Learning Deep RL Robotics Computer Vision

Selected Work

Research publications

Paul Humke^*, and Khang Vo Huynh^*. "A Peano Coincidence". Journal of Mathematical Analysis and Applications, vol. 562 (2026), issue 1.

^* Co first-author. Equal contribution from both authors

Project page

Khang Vo Huynh^*, David Parker, and Lu Feng. "Optimization-Based Robust Controller Synthesis for Interval MDPs". Preprint.

Project page

Paul Humke^*, Khang Vo Huynh^*, and Thong Vo^*. "Efficiently Filling Space". The Rocky Mountain Journal of Mathematics, vol. 53 (2023), no. 2, June 2023, pp. 477-484.

^* Co first-author. Equal contribution from all three authors

Project page

Paul Humke^*, and Khang Vo Huynh^*,. "Finding Keys to the Peano Curve." Acta Mathematica Hungarica, vol. 167, no. 1, 5 July 2022, pp. 255-277.

^* Co first-author. Equal contribution from both authors

Project page