Papers
3
Total Citations
27
H-Index
2
About
Prithvi Poddar is a rising researcher at the forefront of multi-robot systems and reinforcement learning for robotics. His work focuses on enabling autonomous robots to operate reliably in complex, real-world environments, tackling fundamental challenges in multi-agent coordination, sparse reward learning, and adaptive navigation. Poddar’s most impactful contribution is his 2024 paper on multi-agent deep reinforcement learning for persistent monitoring under sensing, communication, and localization constraints (22 citations). This work addresses the critical problem of coordinating heterogeneous robot teams in GPS-denied environments, proposing motion policies that account for real-world limitations. He has also introduced innovative algorithms to overcome sparse reward challenges in continuous control robotics, such as his Heavy-Tailed Stochastic Policy Gradient (HT-PSG) algorithm (2023, 3 citations), which improves learning efficiency in manipulation and navigation tasks. Additionally, his RE-MOVE framework (2023, 2 citations) pioneers the use of language-based feedback to help robotic navigation policies adapt dynamically to changing environments during deployment. Through these contributions, Poddar is advancing the practical deployment of intelligent, resilient robotic systems.
Research Focus
Key Achievements
Top Papers
- 1
- 2
- 3