Papers
15
Total Citations
93
H-Index
6
About
Bharadwaj Amrutur is a leading researcher in robotics and artificial intelligence, focusing on robot manipulation, locomotion, and human-robot interaction. His major contributions span imitation learning for high-precision tasks, such as peg-in-hole insertion, where his work demonstrates generative methods that enable industrial robots to match human-level dexterity (16 citations). He has also advanced quadrupedal locomotion, developing trajectory-based deep policy search and linear policy approaches for robust walking on sloped terrains, with notable work on active spine behaviors for dynamic efficiency. In human-robot collaboration, Amrutur pioneered one-shot object localization using Siamese networks and natural language instruction translation into computer programs for robot manipulation, alongside innovative methods for teaching robots novel objects via pointing. His impact is evident across highly cited papers, including the CORNET 2.0 co-simulation middleware for multi-robot systems (10 citations). Amrutur’s work bridges simulation and real-world deployment, with achievements in low-cost hardware implementation, such as the Stoch 2 quadruped robot, and sensor interrogation systems using fiber Bragg gratings. His research is essential for students and researchers interested in advancing autonomous robotics, imitation learning, and intuitive human-robot interfaces.
Research Focus
Key Achievements
Top Papers
- 1Imitation Learning for High Precision Peg-in-Hole Tasks16 citations · 2020
- 2CORNET 2.0: A Co-Simulation Middleware for Robot Networks10 citations · 2022
- 3One-Shot Object Localization Using Learnt Visual Cues via Siamese Networks10 citations · 2019
- 4
- 5Trajectory based Deep Policy Search for Quadrupedal Walking8 citations · 2019
- 6
- 7
- 8Teaching Robots Novel Objects by Pointing at Them5 citations · 2020
- 9
- 10