Competency-Aware Planning for Probabilistically Safe Navigation Under Perception Uncertainty

TL;DR

This paper introduces PaRCE, a probabilistic method to estimate perception model uncertainty, enhancing safe navigation for unmanned ground vehicles by reducing collisions and improving efficiency in complex terrains.

Contribution

The paper presents a novel competency estimation approach that predicts model familiarity and regional uncertainty, improving navigation safety and efficiency under perception uncertainty.

Findings

Competency scores accurately predict classification correctness and OOD samples.

Regional competency maps distinguish familiar from unfamiliar regions effectively.

Competency-aware navigation reduces collision rates significantly.

Abstract

Perception-based navigation systems are useful for unmanned ground vehicle (UGV) navigation in complex terrains, where traditional depth-based navigation schemes are insufficient. However, these data-driven methods are highly dependent on their training data and can fail in surprising and dramatic ways with little warning. To ensure the safety of the vehicle and the surrounding environment, it is imperative that the navigation system is able to recognize the predictive uncertainty of the perception model and respond safely and effectively in the face of uncertainty. In an effort to enable safe navigation under perception uncertainty, we develop a probabilistic and reconstruction-based competency estimation (PaRCE) method to estimate the model's level of familiarity with an input image as a whole and with specific regions in the image. We find that the overall competency score can correctly predict correctly classified, misclassified, and out-of-distribution (OOD) samples. We also confirm that the regional competency maps can accurately distinguish between familiar and unfamiliar regions across images. We then use this competency information to develop a planning and control scheme that enables effective navigation while maintaining a low probability of error. We find that the competency-aware scheme greatly reduces the number of collisions with unfamiliar obstacles, compared to a baseline controller with no competency awareness. Furthermore, the regional competency information is very valuable in enabling efficient navigation.