Changliu Liu

About

Changliu Liu is a roboticist and AI researcher whose work sits at the intersection of motion planning, human-robot interaction, and safe autonomy. She is perhaps best known for developing the Convex Feasible Set (CFS) algorithm, a groundbreaking approach to real-time motion planning that reframes highly nonconvex optimization problems in cluttered environments into tractable convex ones — a contribution that has garnered over 130 citations and become a cornerstone reference in trajectory optimization. Her broader research agenda tackles one of robotics' most pressing challenges: enabling robots to operate safely and efficiently alongside humans. Liu's work on human-robot collaboration spans algorithmic safety measures for industrial co-robots, robust plan recognition, and sophisticated human motion prediction using adaptable recurrent neural networks and inverse kinematics — tools that allow robots to anticipate human movement and respond intelligently. More recently, she has pushed into humanoid robotics, with her H2O framework enabling real-time whole-body teleoperation using only an RGB camera. Her 2023 survey on state-wise safe reinforcement learning further reflects her commitment to bridging the gap between simulation success and real-world deployment. Collectively, her publications represent a rigorous, safety-first vision for the future of intelligent robotics.

Research Focus

Key Achievements

Top Papers

1. 1
   The Convex Feasible Set Algorithm for Real Time Optimization in Motion Planning

   133 citations · 2018
2. 2
   Towards Efficient Human-Robot Collaboration With Robust Plan Recognition and Trajectory Prediction

   98 citations · 2020
3. 3
   Learning Human-to-Humanoid Real-Time Whole-Body Teleoperation

   68 citations · 2024
4. 4
   Algorithmic safety measures for intelligent industrial co-robots

   62 citations · 2016
5. 5
   Control in a Safe Set: Addressing Safety in Human-Robot Interactions

   57 citations · 2014
6. 6
   Human Motion Prediction using Semi-adaptable Neural Networks

   52 citations · 2019
7. 7
   Human Motion Prediction Using Adaptable Recurrent Neural Networks and Inverse Kinematics

   47 citations · 2020
8. 8
   Convex feasible set algorithm for constrained trajectory smoothing

   43 citations · 2017
9. 9
   State-wise Safe Reinforcement Learning: A Survey

   41 citations · 2023
10. 10
    Real time trajectory optimization for nonlinear robotic systems: Relaxation and convexification

    38 citations · 2017

Key Collaborators

Contact & Links