Reproduce “Prediction uncertainty-aware planning using deep ensembles and trajectory optimisation” 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

Anshul Nayak, Azim Eskandarian

The authors integrate a deep-ensemble pedestrian forecaster with constrained model-predictive control so a robot plans around predicted human motion and its uncertainty. In ETH/UCY simulations, the control-barrier-function planner reached a 16.8 m path and 10.8 s elapsed time, versus 38.1 m and 20.04 s for hard constraints; both succeeded. These findings are simulation-based, not evidence of physical navigation or safety.

Summary by alphaXiv.

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