Shubham Sonawani
Papers
7
Total Citations
22
H-Index
3
About
Shubham Sonawani is a robotics researcher pushing the boundaries of human-robot collaboration and imitation learning. His work centers on making robot control more intuitive, accessible, and ubiquitous—primarily through smartwatch-based interfaces and advanced machine learning. Sonawani’s key contributions include developing **iRoCo**, a framework enabling intuitive robot control from anywhere using a smartwatch and smartphone, and **Diff-Control**, a stateful diffusion-based policy for imitation learning that tackles the challenge of consistent action during robot execution. He has also pioneered methods for human arm pose estimation from a single smartwatch, allowing for anytime, anywhere teleoperation, and explored how visual cues—both static and dynamic—can project robot intentions to improve collaboration. Notably, his work on **SiSCo** leverages Large Language Models to synthesize visual signals for human-robot communication, reducing the need for specialized resources. With multiple papers from 2022-2024 accumulating citations, Sonawani’s research is gaining traction for its practical, real-world impact. His modular, attention-based approach to language-conditioned policies further demonstrates his commitment to efficient, transferable robot learning. Sonawani is shaping a future where robots are seamlessly integrated into daily life, controlled naturally and intuitively from anywhere.
Research Focus
Key Achievements
Top Papers
- 1Diff-Control: A Stateful Diffusion-based Policy for Imitation Learning5 citations · 2024
- 2
- 3
- 4iRoCo: Intuitive Robot Control From Anywhere Using a Smartwatch2 citations · 2024
- 5
- 6Learning modular language-conditioned robot policies through attention2 citations · 2023
- 7