Reproduce “FASTER: Fast and Safe Trajectory Planner for Navigation in Unknown Environments” 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

Jesus Tordesillas, Brett T. Lopez, Michael Everett, Jonathan P. How

The FASTER framework enables high-speed, safe navigation for UAVs and ground robots in unknown environments by concurrently planning an aggressive "Whole Trajectory" that extends into unknown space and a conservative "Safe Trajectory" solely within known free space for guaranteed fallback. This dual-trajectory approach, coupled with efficient optimization, demonstrated up to a 52.3% improvement in average completion time in simulations and achieved real-world UAV flight speeds of 7.8 m/s.

Key insights

  • Simultaneously planning an ambitious trajectory into unknown space and maintaining a conservative, safe backup trajectory within known free space effectively resolves the speed-safety tradeoff.
  • Optimizing trajectory interval allocation via MIQP and adaptively determining time per interval significantly reduces conservatism and improves trajectory efficiency compared to fixed or ad-hoc schemes.
  • The 'Safe Trajectory' is crucial for high-speed collision avoidance, providing a guaranteed stopping mechanism when new obstacles are rapidly detected.

Summary by alphaXiv.

Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.