A Robust Sidewalk Navigation Method for Mobile Robots Based on Sparse Semantic Point Cloud

Mingxing Wen,Yunxiang Dai,Tairan Chen,Chunyang Zhao,Jun Zhang,Danwei W. Wang

DOI: https://doi.org/10.1109/IROS47612.2022.9981281

2022-10-23

Abstract:Last-mile delivery robots are usually required to navigate on the sidewalk through a fixed route. The current solutions heavily rely on the image-based perception and GPS localization to successfully complete delivery tasks. However, it is prone to fail and become unreliable when the robot runs in challenging conditions, such as operating in different illuminations, or under canopies of trees or buildings. To address these issues, this paper proposes a novel robust sidewalk navigation method for the last-mile delivery robots with an affordable sparse LiDAR, which consists of two main modules: Semantic Point Cloud Network (SegPCn) and Reactive Nav-igation Network (RNn), as shown in Fig. 1. More specifically, SegPCn takes the raw 3D point cloud as input and predicts the point-wise segmentation labels, presenting a robust perception capability even in the night. Then, the semantic point clouds are fed to RNn to generate an angular velocity to navigate the robot along the sidewalk, where the localization of the robot is not required. Moreover, an autolabeling mechanism is developed to reduce the labor involved in data preparation as well. And the LSTM neural network is explored to effectively leverage the historical context and derive correct decisions. Extensive experiments have been carried out to verify the efficacy of this method, and the results show that this method enables the robot to navigate on the sidewalk robustly during day and night. We open source the code and the data set on https://github.com/lukewenMX/Robust-Navigation-Method.

Computer Science