Joint Localization and Planning using Diffusion

TL;DR

This paper introduces a diffusion-based model for joint localization and path planning in 2D environments, enabling real-time navigation by integrating perception and planning from egocentric LIDAR data.

Contribution

It presents a novel diffusion model that generates collision-free paths conditioned on obstacles and sensor data, extending diffusion applications to end-to-end robotic navigation.

Findings

Model generalizes to different map appearances

Accurately describes ambiguous localization solutions

Operates in real-time for end-to-end navigation

Abstract

Diffusion models have been successfully applied to robotics problems such as manipulation and vehicle path planning. In this work, we explore their application to end-to-end navigation -- including both perception and planning -- by considering the problem of jointly performing global localization and path planning in known but arbitrary 2D environments. In particular, we introduce a diffusion model which produces collision-free paths in a global reference frame given an egocentric LIDAR scan, an arbitrary map, and a desired goal position. To this end, we implement diffusion in the space of paths in SE(2), and describe how to condition the denoising process on both obstacles and sensor observations. In our evaluation, we show that the proposed conditioning techniques enable generalization to realistic maps of considerably different appearance than the training environment, demonstrate our model's ability to accurately describe ambiguous solutions, and run extensive simulation experiments showcasing our model's use as a real-time, end-to-end localization and planning stack.