Papers
6
Total Citations
54
H-Index
5
About
Andrew Clark is a researcher whose work sits at the intersection of cyber-physical systems (CPS) security, control theory, and reinforcement learning, with a particular focus on ensuring the safety and resilience of autonomous systems operating in adversarial or fault-prone environments. His most influential contribution lies in the application of Control Barrier Functions (CBFs) to guarantee system safety under real-world threats, including sensor attacks, actuator failures, and environmental uncertainty. His 2020 paper on CBFs for safe CPS under sensor faults and attacks, which has garnered 14 citations, exemplifies his approach of bridging rigorous mathematical control frameworks with practical security concerns. Clark has also made notable strides in safe reinforcement learning, combining model-based RL with provable safety guarantees through CBFs, addressing one of the field's most challenging open problems. His work on interactive reward shaping via human feedback demonstrates a broader interest in human-AI collaboration within complex learning environments. More recently, his timing-based resilience framework for CPS extends his contributions to real-world critical infrastructure domains including robotics, manufacturing, and power systems. Across his portfolio, Clark consistently advances the theoretical foundations needed to deploy autonomous systems safely and reliably in high-stakes environments.
Research Focus
Key Achievements
Top Papers
- 1Control Barrier Functions for Safe CPS Under Sensor Faults and Attacks14 citations · 2020
- 2
- 3
- 4
- 5
- 6Resilient Trajectory Planning in Adversarial Environments2 citations · 2019