Reproduce “BIG-CBF: Behavior-Imagination-Guided Control Barrier Function with Shared Uncertainty for Mobile Robot Navigation” on OpenResearch
Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.
Universal app for Apple silicon and Intel. Requires macOS 11 or later.
About this paper
Shibo Li, Zhongcheng Wang, Jiahe Cao, Jianhua Yang, Ke Wu
BIG-CBF augments a hard control barrier function with short-horizon imagination over six closed-loop behaviors, allowing a mobile robot to choose persistent passing, sidestepping, or yielding maneuvers while retaining high-rate safety filtering. Across 3,600 simulated episodes it achieved 99.78% success with zero collisions and, on a physical robot, completed all 15 evaluated trials while shared uncertainty reduced CBF intervention and activation.
Key insights
- Behavior imagination addresses CBF liveness failures by evaluating whether a candidate maneuver both remains barrier-compatible and makes progress toward the goal.
- Using the same uncertainty sources in planning and execution substantially reduces downstream correction: in simulation, intervention energy fell from 0.10 to 0.02 and CBF activation from 15.64% to 9.10% relative to the no-shared-uncertainty ablation.
- The shared uncertainty model improves robustness at the cost of modest conservatism, increasing simulated completion time from 11.21 to 11.72 seconds.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.