PIETRA: Physics-Informed Evidential Learning for Traversing Out-of-Distribution Terrain

TL;DR

PIETRA is a self-supervised learning framework that combines physics priors with evidential neural networks to improve off-road traversability prediction, especially in out-of-distribution terrains, enhancing navigation safety and accuracy.

Contribution

It introduces a novel physics-informed evidential learning method that seamlessly integrates physics knowledge into neural networks for better out-of-distribution terrain handling.

Findings

Improves traversability prediction accuracy in diverse terrains.

Enhances navigation performance under distribution shifts.

Demonstrates effectiveness through simulations and hardware tests.

Abstract

Self-supervised learning is a powerful approach for developing traversability models for off-road navigation, but these models often struggle with inputs unseen during training. Existing methods utilize techniques like evidential deep learning to quantify model uncertainty, helping to identify and avoid out-of-distribution terrain. However, always avoiding out-of-distribution terrain can be overly conservative, e.g., when novel terrain can be effectively analyzed using a physics-based model. To overcome this challenge, we introduce Physics-Informed Evidential Traversability (PIETRA), a self-supervised learning framework that integrates physics priors directly into the mathematical formulation of evidential neural networks and introduces physics knowledge implicitly through an uncertainty-aware, physics-informed training loss. Our evidential network seamlessly transitions between learned and physics-based predictions for out-of-distribution inputs. Additionally, the physics-informed loss regularizes the learned model, ensuring better alignment with the physics model. Extensive simulations and hardware experiments demonstrate that PIETRA improves both learning accuracy and navigation performance in environments with significant distribution shifts.