Papers
15
Total Citations
1,077
H-Index
12
About
X Xinjilefu is a leading figure in humanoid robotics, best known for pioneering optimization-based full-body control that enabled the Atlas robot to achieve unprecedented reliability. As a key member of Team WPI-CMU during the DARPA Robotics Challenge (DRC), Xinjilefu developed a cascaded optimization framework that replaced traditional inverse kinematics with real-time inverse dynamics. This approach, detailed in papers with over 229 and 163 citations, provided compliant, robust motions and allowed Atlas to perform complex disaster-response tasks without falling or requiring physical resets—a landmark achievement in the field. Their work on online footstep optimization and center-of-mass estimation further advanced dynamic walking and fall prevention, with contributions cited over 1,000 times collectively. Xinjilefu’s controllers were instrumental in the DRC Finals, where their "no falls, no resets" strategy proved that humanoid robots could operate reliably in unstructured environments. For students and researchers, Xinjilefu’s research represents a gold standard in marrying theoretical optimization with practical, real-time control, demonstrating how rigorous algorithmic design can transform a robot from a fragile prototype into a capable, resilient platform for disaster response and beyond.
Research Focus
Key Achievements
Top Papers
- 1Optimization‐based Full Body Control for the DARPA Robotics Challenge229 citations · 2015
- 2Optimization based full body control for the atlas robot163 citations · 2014
- 3
- 4No falls, no resets: Reliable humanoid behavior in the DARPA robotics challenge105 citations · 2015
- 5Robust dynamic walking using online foot step optimization91 citations · 2016
- 63D walking based on online optimization83 citations · 2013
- 7
- 8What Happened at the DARPA Robotics Challenge Finals57 citations · 2018
- 9
- 10