Shalabh Bhatnagar
Papers
7
Total Citations
47
H-Index
4
About
Shalabh Bhatnagar is a robotics and artificial intelligence researcher whose work sits at the intersection of deep reinforcement learning, autonomous navigation, and legged locomotion. His research has made meaningful contributions to two distinct but complementary domains: UAV obstacle avoidance and quadrupedal robot locomotion. In the area of autonomous aerial systems, Bhatnagar developed memory-based deep reinforcement learning frameworks enabling UAV quadrotors equipped with monocular cameras to navigate complex, unknown indoor environments — work that has garnered over 16 citations and addressed a particularly challenging problem given the dynamic instability of aerial platforms compared to ground vehicles. His most sustained contributions lie in quadrupedal locomotion, where he has pioneered trajectory-based deep policy search methods, linear policy approaches for sloped terrain navigation, and kinematic motion primitive frameworks for the quadruped robot Stoch 2. Notably, his gait library synthesis work using augmented random search demonstrated practical deployment of learned behaviors — including forward trot, backward trot, and turning gaits — on low-cost hardware, bridging the gap between simulation and real-world robotics. Across seven key publications, Bhatnagar's research collectively advances the field of embodied AI, offering scalable, deployable solutions for robots operating in unstructured real-world environments.
Research Focus
Key Achievements
Top Papers
- 1
- 2Trajectory based Deep Policy Search for Quadrupedal Walking8 citations · 2019
- 3
- 4
- 5
- 6Gait Library Synthesis for Quadruped Robots via Augmented Random Search4 citations · 2019
- 7