Chenggang Liu

Humanoid Robot Full-body Motion Planning and Control

Carnegie Mellon University, Pittsburgh, 2013-2015

Description

Developed full-body motion planning and control for the Atlas humanoid robot's car egress task in the DARPA Robotics Challenge. The team achieved zero falls and zero resets during the competition through robust whole-body control and careful system engineering. (Slides)

Publications

  • Team WPI-CMU: Achieving Reliable Humanoid Behavior in the DARPA Robotics Challenge. Mathew DeDonato, Felipe Polido, Kevin Knoedler, et al. Journal of Field Robotics 34(2), 381-399
  • What Happened in the DARPA Robotics Challenge, and Why? C. G. Atkeson, et al. Journal of Field Robotics
  • Full-body motion planning and control for the car egress task of the DARPA Robotics Challenge. Chenggang Liu, Christopher G Atkeson, Siyuan Feng, X Xinjilefu. Humanoids 2015
  • No falls, no resets: Reliable humanoid behavior in the DARPA Robotics Challenge. Christopher G Atkeson, et al. Humanoids 2015

Demos

Video demo for Humanoids 2015

Simulation result without using tools

Monte Carlo simulation for finding optimal egress strategies

Inverse-kinematics solver for interactive pose setting 1

Inverse-kinematics solver for interactive pose setting 2

Teleoperation/Monitoring UI

DRC Final garage rehearsal

Rough terrain walking

This video demonstrates how challenging it is to let a robot get out of the car by itself

Team Trooper egress planning

Team TRAC egress planning

Team Tartan egress planning

Team MIT egress planning

Team NIMBRO egress planning

Team IHMC egress planning

Team KAIST egress planning