SLIP based Human-Inspired Control applied to the compliant underactuated bipedal robot ATRIAS. Using bio-inspired virtual constraints and canonical walking functions, the full-order robot can be reduced to a low-order model. Walking gaits are generated through the use of a novel multi-domain optimization that leverages this low-dimensional representation. These gaits are verified in simulation, and implemented on the ATRIAS. The end result is remarkably natural looking walking.
More information at: www.bipedalrobotics.com
Special thanks to:
Jonathan Hurst and the Dynamic Robotics Laboratory: / oregonstatedrl
Jessy Grizzle and DynamicLegLab: / dynamicleglocomotion