Our lab focuses in addressing planning and control problems centering around legged robots. Currently, we are working on the following topics:
- Legged mobility in complex environments, e.g., sand and snow: how we can bring together rigorous model-based approaches with advanced efficient data-driven methods to realize robust and adaptive locomotion behaviors in complex environments.
- Dynamic motion synthesis: are our current approaches sufficient for realizing highly-dynamic behaviors on complex robots efficiently, robustly, and safely?
- Design, planning, and control that best utilize the dynamics structures of legged robots: Legged robots have dynamics that are particularly different than these of drones or cars. How do we best utilize their dynamics structures of legged robots that constantly make and break contact with the environment? Moreover, how can we exploit the second order Lagrangian dynamics of robotic systems in general? Machine learning is being proven powerful in robotics; yet the optimal combinations of using known models and exploiting data online using their dynamics structures are to be synthesized.
