Sahil Badyal
Papers
2
Total Citations
8
H-Index
2
About
Sahil Badyal’s research lies at the intersection of reinforcement learning, multiagent systems, and partially observable Markov decision processes (POMDPs), with a strong emphasis on real-world autonomous decision-making. His major contributions include developing novel multiagent rollout and policy iteration algorithms that enable coordinated sequential repair and maintenance tasks in robotics, even under partial observability. In his most-cited work (2020, 6 citations), Badyal introduced a framework that simultaneously or sequentially optimizes agents’ controls using multistep lookahead and truncated rollout, significantly improving efficiency in multi-robot repair problems. A subsequent paper (2020, 2 citations) refined these ideas for single-agent autonomous repair, demonstrating how a known base policy and terminal cost approximation can yield near-optimal performance in infinite-horizon, discounted settings. Though early in his career, Badyal’s work is notable for bridging theoretical dynamic programming with practical, scalable algorithms—a critical step toward deploying autonomous systems in logistics, infrastructure maintenance, and disaster response. His research is particularly valuable for students and engineers seeking to understand how reinforcement learning can be made tractable in complex, partially observable environments.
Research Focus
Key Achievements
Top Papers
- 1
- 2