- Bharadwaj Amrutur (RBCCPS/Electrical Communication Engineering)
- Shalabh Bhatnagar (RBCCPS/Computer Science and Automation)
- Ashitava Ghosal (Mechanical Engineering)
- Shishir N. Y. Kolathaya (RBCCPS)
RBCCPS has a custom built four legged robot (quadruped) called “Stoch”. The main goal is to explore deep-reinforcement learning algorithms to achieve a wide variety of walking behaviours. The quadruped must be able to plan and navigate over a diverse set of environments like flat plains, undulating and slippery surfaces, and also muddy terrains. Besides walking, we are also interested in exploring applications like:
- The quadruped must use vision, inertial measurements and other sensory information to read its surroundings, and then plan and execute its motion to a designated point.
- A traditional arm mechanism attached to the quadruped must be able to pick and place movable objects from one point to another.
STOCH in action
Mathematical model for pressure–deformation relationship of miniaturized McKibben actuators Conference
Proceedings of the 3rd International and 18th National Conference on Machines and Mechanisms (iNaCoMM), 13.-15.12.17, Mumbai, 2018.
Realizing learned quadruped locomotion behaviors through kinematic motion primitives Journal Article
arXiv: Computer Science, 2018.
Trajectory planning and obstacle avoidance for hyper-redundant serial robots Journal Article
Journal of Mechanisms and Robotics, 9 (4), pp. 041010:1-8, 2017.