Chao Cao

I am a first-year Ph.D. student at The Robotics Institute, Carnegie Mellon University advised by Ji Zhang and Howie Choset.

I serve as the Planning Lead of Team Explorer competing in the DARPA Subterranean Challenge. For a teaser, check out this cool video from Microsoft and the compilation of our Urban Circuit performance !

I have an MS in Robotics from CMU RI, where I was advised by Matt Travers and Howie Choset. I obtained my BS in Computer Science from The Univeristy of Hong Kong, where I worked with Jia Pan and Wenping Wang.

I'm interested in robot navigation and motion planning. My recent work has been on exploration and sensor coverage planning.

Home / Robots / Google Scholar / CV

News

Jul 2021 : Paper on Autonomous Exploration won the Best Paper Award and Best Systems Paper Award of RSS 2021.
Feb 2020 : Came in second place in the SubT Urban Circuit.
Jan 2020 : Paper on Coverage Path Planning accepted to ICRA 2020.
Aug 2019 : Came in first place in the SubT Tunnel Circuit.
Jan 2019 : Paper on Crowd Navigation accepted to ICRA 2019.

Research

Efficient Autonomous Exploration

DARPA Subterranean Challenge, 2020
video /

A hierarchical framework for autonomous exploration in large-scale environments.

Hierarchical Coverage Path Planning in Complex 3D Environments

Chao Cao, Ji Zhang, Matt Travers and Howie Choset
IEEE International Conference on Robotics and Automation (ICRA), 2020
video /

A multi-resolution hierarchical framework for sensor coverage planning. It solves a high-level Traveling Salesman Problem (TSP) for a global tour, which is then used for assembling low-level trajectories obtained by solving local Orienteering Problems. The hierarchical scheme produces much higher efficiency than the state-of-the-art.

Dynamic Channel: A Planning Framework for Crowd Navigation

Chao Cao, Peter Trautman and Soshi Iba
IEEE International Conference on Robotics and Automation (ICRA), 2019
arxiv / video /

A hierarchical planning framework for crowd navigation. It incorporates a global planner to find feasible paths in the topological space efficiently, and a local planner to follow the path with optimized trajectories.

Who Visited Me

Design and source code from Leonid Keselman's website