Papers
5
Total Citations
19
H-Index
3
About
Kushal Kedia is a rising researcher at the intersection of human-robot interaction, motion planning, and game-theoretic decision-making. His work addresses a fundamental challenge in robotics: enabling robots to anticipate and adapt to human behavior in real-time collaborative settings. In his highly cited paper "InteRACT," Kedia introduces transformer models that solve the chicken-or-egg problem of human intent prediction conditioned on robot actions—a breakthrough for fluid human-robot manipulation. His game-theoretic framework for joint forecasting and planning tackles the critical issue of long-tail safety in human-robot environments, moving beyond simple most-likely predictions. Kedia has also advanced kinodynamic planning for vehicular systems navigating uneven terrains, developing terrain-aware learned controllers that dramatically improve planning efficiency. His work on sampling-based motion planning using learned critical sources addresses the practical challenge of finding feasible paths in reasonable time. With papers spanning from 2020 to 2025, including his latest work on one-shot imitation under mismatched execution, Kedia's research consistently pushes toward more capable, safe, and intuitive robotic systems that can work alongside humans in complex, unstructured environments.
Research Focus
Key Achievements
Top Papers
- 1
- 2
- 3A Game-Theoretic Framework for Joint Forecasting and Planning4 citations · 2023
- 4
- 5One-Shot Imitation Under Mismatched Execution1 citations · 2025