Autonomous Vehicle Challenges for the US Rural Midwest

Principal Investigator(s):

Rajesh Rajamani, Professor, Chair, Mechanical Engineering

Project summary:

While major global companies such as Google, Uber, Tesla and General Motors are intensively focused on developing Level 3 and higher self-driving autonomous vehicles, their research is focused on driving in good weather on clean well-equipped roads. A majority of their development and testing has occurred in warm weather states such as California and Arizona. States in the US Midwest and rural locations pose additional challenges for autonomous driving that could significantly delay the arrival of self-driving technology to these regions. The use of cameras is the most common approach for determining a vehicle's position in the lane and measuring its lateral distance to the lane markers. This lateral distance to lane markers is used as the feedback variable for automatic steering control. The presence of snow on the ground, including even a thin layer of frost covering the lane markers, makes the engagement of steering control infeasible. Data gathered by the University of Minnesota team shows that snow remains on lane markers for a significantly longer duration, compared to snow in the interior of the lane where the tire-paths of cars enable faster snow removal. This prevents the engagement of autonomous steering for a significant period of time after the end of snowfall. Further, low volume rural roads present special challenges in all seasons. Such rural roads can be narrow, often will not have right-side lane markers, may not even have center lines, may not be plowed for snow removal, and can have trees and other objects close to the side of the road. Rural traffic intersections can have missing delineation and signage that are normally provided on higher volume roadways. All of these issues pose major challenges to autonomous driving, since AVs depend critically on such markings and signage. This project will study the challenges for autonomous vehicles in the Midwest, including winter driving in cities and all-season driving in rural areas. It will evaluate potential solutions to these challenges.

Project details: