Feature mapping and state estimation for highly automated vehicles

Abstract

The past decade has witnessed an acceleration of autonomous vehicle research and development, with technological advances contributed by academia, government, and the industrial and consumer sectors. These advancements hold the potential to improve society by enhancing transportation safety and throughput, where decreased congestion saves time and reduces vehicle emissions. Two of the key technologies to enable vehicle infrastructure interaction, advanced traffic management, and automated vehicles are automated roadway mapping and reliable vehicle state estimation. In this paper, we present an overview and new methods for the problems automated roadway mapping plus a discussion of the extension of these methods to the problem of vehicle state estimation. Results from the application of these methods to feature mapping and state estimation are presented.

Keywords:

• automated precise mapping
• vehicle state estimation
• Bayesian estimation
• vehicle infrastructure interaction
• advanced driver assistance systems
• autonomous driving

Acknowledgements

The authors also acknowledge many of the collaborators who made technical contributions in this project including Matthew Barth, Anning Chen, Arvind Ramanandan, Qichi Yang, and Sheng Zhao.

Notes

1 Note that the attitude corrections are not additive. Nonetheless, this equation conveys the idea of the corrective update.

Additional information

Funding

This work was supported in part by the US Department of Transportation Federal Highway Administration [grant number DTFH61-09-C-00018] and [grant number DTFH61-06-D-00006], and California Department of Transportation [grant number 65A0261].