Non-Vector Event-Based Method for Mobile Robot Navigation in Narrow Environment

Abstract

<p>Wheeled mobile robots have been exten-sively deployed across various domains, particularly in the context of service robots for indoor applications. However, navigating through narrow indoor environments has significant challenges since robots relying on SLAM (Simultaneous Localization and Mapping) algorithms may fail to extract sufficient features, leading to substantial localization errors or even failures, thereby impeding the completion of navigation tasks. To address this issue, we propose an event-based non-vector space navigation framework specifically designed for narrow space navigation. Firstly, we introduce a non-vector space localization method based on the event-based theory to calculate the current state of the robot. Subsequently, we present a non-vector space controller aimed at executing tracking tasks. We have tested the proposed algorithm in actual narrow environments, where the robot successfully and stability navigated in a narrow restroom.<br></p>