Legged robots are accessible to unstructured environments, which enables legged robots to aid in package delivery, terrain exploration, search and rescue, and disaster relief, and becoming assistants in our homes. Two-legged robots (i.e., bipedal robots) with tall and slim shapes can easily adopt to structures built for humans (narrow staircases or passages). Besides assisting the elderly or people with physical disabilities, research on bipedal robots and exoskeletons habilitates humans to improve productivity. Moreover, the knowledge among bipedal robots, exoskeletons, prosthetics, and humans is transferable and becomes positive reinforcement. These merits motivate me to work on autonomy of bipedal robots.

The capabilities of bipedal robots have yet to be unleashed to serve the society due to a number of challenges. One key challenge involves problems in autonomy system integration and smooth motion planning, as they are critical for a bipedal robot to autonomously walk toward a distant destination and to smoothly avoid obstacles detected from different calibrated sensors while maintaining its stability.
 

In particular, I seek to develop a full autonomy system that allows bipedal robots to

1) Acquire multi-modal data from a calibrated perception suite;

2) Estimate their poses in textureless environments;

3) Detect and avoid dynamic obstacles;

4) Traverse unexplored, unstructured environments and undulating terrains.

5) Perform point-to-point topometric navigation.

All the research conducted focuses on advancing the state-of-the-art algorithms that will achieve autonomy of bipedal robots — Cassie Blue and Digit.