GNSS Hardware in the Loop Simulation for Autonomous Racing Localization Development. Karlins, B. J., Harris, J. W., Meyer, S. W., & Bevly, D. M. In pages 391–401, April, 2024.
GNSS Hardware in the Loop Simulation for Autonomous Racing Localization Development [link]Paper  doi  abstract   bibtex   
Development of any autonomous mobile robot platform will invariably involve various stages of testing individual components, and the full integrated platform. For many autonomous vehicles, a combination of operating cost, platform availability, operational environment, or other logistical constraints makes full scale live testing impractical. Furthermore, even when live testing is possible, test requirements such as high repeatability or introduction of edge cases may not be achievable. Autonomous Racing is a robotics field that acutely struggles from these testing constraints, and as such, various bespoke simulation tools have been developed by research teams studying this area. GNSS based localization algorithms are critical to safe and reliable operation of these vehicles, but are limited in their ability to be tested in simulation due to the difficulty of modeling intricate receiver behaviors in complex scenarios. Component behavior modeling for simulation development can be bypassed by directly including the test hardware in the simulation loop, also known as a Hardware-in-the-Loop simulation. This paper covers the initial design and testing of a GNSS Hardware-in-the-Loop simulation system for autonomous vehicles with a particular focus on autonomous racing. The simulation system is designed as an extension to existing software-in-the-loop simulation tools developed by Autonomous Tiger Racing, a research group competing in the Indy Autonomous Challenge. Preliminary tests and experiments are performed on this nascent system, and current capabilities and outstanding issues are identified.
@inproceedings{karlins_gnss_2024,
	title = {{GNSS} {Hardware} in the {Loop} {Simulation} for {Autonomous} {Racing} {Localization} {Development}},
	url = {http://www.ion.org/publications/abstract.cfm?jp=p&articleID=19644},
	doi = {10.33012/2024.19644},
	abstract = {Development of any autonomous mobile robot platform will invariably involve various stages of testing individual components, and the full integrated platform. For many autonomous vehicles, a combination of operating cost, platform availability, operational environment, or other logistical constraints makes full scale live testing impractical. Furthermore, even when live testing is possible, test requirements such as high repeatability or introduction of edge cases may not be achievable. Autonomous Racing is a robotics field that acutely struggles from these testing constraints, and as such, various bespoke simulation tools have been developed by research teams studying this area. GNSS based localization algorithms are critical to safe and reliable operation of these vehicles, but are limited in their ability to be tested in simulation due to the difficulty of modeling intricate receiver behaviors in complex scenarios. Component behavior modeling for simulation development can be bypassed by directly including the test hardware in the simulation loop, also known as a Hardware-in-the-Loop simulation. This paper covers the initial design and testing of a GNSS Hardware-in-the-Loop simulation system for autonomous vehicles with a particular focus on autonomous racing. The simulation system is designed as an extension to existing software-in-the-loop simulation tools developed by Autonomous Tiger Racing, a research group competing in the Indy Autonomous Challenge. Preliminary tests and experiments are performed on this nascent system, and current capabilities and outstanding issues are identified.},
	language = {en},
	urldate = {2024-06-20},
	author = {Karlins, Bryce J. and Harris, Jacob W. and Meyer, Stephanie W. and Bevly, David M.},
	month = apr,
	year = {2024},
	pages = {391--401},
}

Downloads: 0