Radha Poovendran

University of Washington

Papers

3

Total Citations

26

H-Index

3

About

Radha Poovendran is a leading researcher at the intersection of reinforcement learning, cyber-physical systems (CPS), and safe robotics. Their work focuses on ensuring that autonomous agents can operate reliably under real-world constraints, including human interaction, sensor faults, and cyber attacks. A key contribution is the development of "FRESH," an interactive reward-shaping framework that leverages human feedback to guide reinforcement learning agents in high-dimensional state spaces—a critical step toward making AI training more efficient and aligned with human intent. Poovendran has also pioneered safety-guarantee mechanisms for CPS, such as a timing-based framework that enforces safety constraints even under faults or attacks, and fault-tolerant neural control barrier functions that protect robotic systems when sensors are compromised. With over 26 citations across their most-cited works, Poovendran’s research is gaining traction for its practical relevance to autonomous driving, industrial automation, and defense. Their recent 2024 paper on neural CBFs showcases a novel fusion of machine learning and formal safety verification, positioning them as a rising authority in resilient, human-aware autonomy.

Research Focus

Key Achievements

3
H-Index
3
Papers
26
Total Citations
9
Avg Citations/Paper
🏆 Most Cited Paper
FRESH: Interactive Reward Shaping in High-Dimensional State Spaces using Human Feedback
13 citations · 2020
📈 Most Prolific Year: 2020 (1 Papers)
🤝 Key Collaborators: 10
🏛 Institutions: University of Washington

Top Papers

  1. 1
  2. 2
  3. 3

Key Collaborators

Contact & Links

Available for collaboration
Content generated · 12 days ago