ColNav: Uncertainty-Aware Spacetime Collision Checking with Bounding Volume Hierarchies
ColNav: Uncertainty-Aware Spacetime Collision Checking with Bounding Volume Hierarchies
This paper introduces ColNav, a robust collision checking method for mobile robots that encodes uncertainty into spacetime tubes organized via bounding volume hierarchies. The approach handles both bounded-set and Gaussian uncertainty descriptions, unifies static and dynamic obstacles, and achieves a 100% success rate under prediction model mismatch compared to 65% for the best baseline while maintaining up to three times the clearance in dense traffic; real-time operation on a physical robot is also validated.
The authors present ColNav, a method that encodes position and velocity uncertainty into spacetime tubes organized via bounding volume hierarchies to achieve robust collision checking; they report a 100% success rate under prediction model mismatch versus 65% for the best baseline in their evaluation.
Source: arXiv Robotics — research abstracts · Read original article ↗
Loading article text…
What the source reports
Publisher-reported claims, with original evidence. These results have not been independently verified by RoboSignal.
Reported numbers
Reported success rate
100%
View original evidence
ers the entire horizon, rather than testing sampled time instants, and returns a robust geometric certificate together with an estimate of the time to collision. Static and dynamic obstacles share the same representation. Our approach reaches the goal with a 100% success rate under prediction model mismatch, where the best baseline reaches 65%, and maintains up to three times the clearance in dense traffic. We also v
Open source S4
- paper: not_reported
What remains unknown
Not established in the collected evidence: Environment, Control, Data origin.
Reported performance applies to the described task. It does not establish general autonomy or deployment readiness.
Source excerpts and review record
Automatically extracted; no manual editorial approval recorded.
alidate real-time operation on a physical robot.
Open source S5
ers the entire horizon, rather than testing sampled time instants, and returns a robust geometric certificate together with an estimate of the time to collision. Static and dynamic obstacles share the same representation. Our approach reaches the goal with a 100% success rate under prediction model mismatch, where the best baseline reaches 65%, and maintains up to three times the clearance in dense traffic. We also v
Open source S4
Source：arXiv Robotics — research abstracts · arxiv.org