Use a Single-Antenna Global Navigation Satellite System (GNSS) to Improve Automated Vehicle Accuracy While Reducing Hardware Costs on Rural Roads.
Minnesota Study Tested Solutions for Automated Vehicle Accuracy on Rural Roads where Vehicle Radars May be Limited.
Minnesota, United States
Autonomous Vehicle Challenges for the US Rural Midwest
Summary Information
Automated vehicles (AVs) have the potential to improve rural road safety, but rural roads present special challenges for AVs (e.g., lack of center lines and signage, unplowed after snow). This study explored non-camera AV solutions for rural driving, testing the availability and accuracy of GNSS for steering control in rural, suburban, and urban locations surrounding Minneapolis, Minnesota. A real-time kinematic (RTK) correction system was used to increase GNSS accuracy. The University of Minnesota’s AV, known as the MnCAV vehicle, was used for all data collection, which took place over the course of the study from December 2023 to August 2025.
Some key considerations from the study included the following:
- Use a single-antenna GNSS to improve AV accuracy while reducing hardware costs. For rural roads without dense tree cover, the low-cost single-antenna solution worked reliably, as compared to a more expensive dual-antenna.
- Consider the use of RTK-corrected GNSS to enhance accuracy. RTK-corrected GNSS achieved localization accuracy with an error standard deviation of less than 10 cm on rural roads with minimal tree obstruction.
- Use other methods to complement GPS-based navigation in downtown environments with tall buildings, narrow streets, and tree cover, all which frequently block or degrade satellite signals. This study tested a Lidar algorithm (LIO-SAM) for these types of GPS outages.
