Estimation and Navigation Methods with Limited Information for Autonomous Urban Driving

TL;DR

This paper explores algorithms and architecture for autonomous urban driving using limited information, addressing challenges like unreliable GPS signals and the need for detailed maps, to enable effective navigation in complex city environments.

Contribution

It introduces novel methods and an architecture for autonomous driving that operate effectively with limited environmental information, reducing reliance on detailed pre-mapped data.

Findings

Algorithms successfully navigate urban environments with limited data

Architecture maintains robust performance despite unreliable localization signals

Approach reduces dependence on extensive prior mapping and data collection

Abstract

Urban environments offer a challenging scenario for autonomous driving. Globally localizing information, such as a GPS signal, can be unreliable due to signal shadowing and multipath errors. Detailed a priori maps of the environment with sufficient information for autonomous navigation typically require driving the area multiple times to collect large amounts of data, substantial post-processing on that data to obtain the map, and then maintaining updates on the map as the environment changes. This dissertation addresses the issue of autonomous driving in an urban environment by investigating algorithms and an architecture to enable fully functional autonomous driving with limited information.