HALO: Human Preference Aligned Offline Reward Learning for Robot Navigation

TL;DR

HALO is a new offline reward learning method that captures human navigation preferences to improve robot navigation, demonstrating superior real-world performance and generalization across environments and hardware.

Contribution

HALO introduces a novel offline reward learning algorithm that aligns robot navigation with human preferences using preference ranking and binary feedback.

Findings

HALO outperforms state-of-the-art methods in success rate and trajectory metrics.

Policies trained with HALO generalize well to unseen environments.

HALO is effective in both learning-based and classical navigation frameworks.

Abstract

In this paper, we introduce HALO, a novel Offline Reward Learning algorithm that quantifies human intuition in navigation into a vision-based reward function for robot navigation. HALO learns a reward model from offline data, leveraging expert trajectories collected from mobile robots. During training, actions are uniformly sampled around a reference action and ranked using preference scores derived from a Boltzmann distribution centered on the preferred action, and shaped based on binary user feedback to intuitive navigation queries. The reward model is trained via the Plackett-Luce loss to align with these ranked preferences. To demonstrate the effectiveness of HALO, we deploy its reward model in two downstream applications: (i) an offline learned policy trained directly on the HALO-derived rewards, and (ii) a model-predictive-control (MPC) based planner that incorporates the HALO reward as an additional cost term. This showcases the versatility of HALO across both learning-based and classical navigation frameworks. Our real-world deployments on a Clearpath Husky across diverse scenarios demonstrate that policies trained with HALO generalize effectively to unseen environments and hardware setups not present in the training data. HALO outperforms state-of-the-art vision-based navigation methods, achieving at least a 33.3% improvement in success rate, a 12.9% reduction in normalized trajectory length, and a 26.6% reduction in Frechet distance compared to human expert trajectories.