Vehicle Automation & Weather: Challenges and Opportunities

Adverse weather has major impacts on the safety and operations of all roads, from signalized arterials to Interstate highways. Weather affects driver behavior, vehicle performance, pavement friction, and roadway infrastructure, thereby increasing the risk of crashes. Weather can be classified as general weather or road weather. General weather relates to measures such as air temperature, precipitation type, fog, dust, etc. while road weather relates to pavement conditions such as surface temperature, surface condition, friction coefficient, etc. Most publications have focused on weather-related collision risks, traffic volume variations, signal control, travel patterns, and traffic flow parameters. However, there is limited literature that characterize vehicle performance at different levels of automation under adverse weather conditions.

Recently, vehicle manufacturers have introduced several automated features such as: adaptive cruise control, automatic braking and lane keeping, park assist, and traffic jam assist. These applications primarily rely on a family of sensors and camera systems. Adverse weather conditions such as fog, heavy rain, snow, and wind can severely limit the functionality and performance of sensors and cameras. Although the applications appear to work well in dry, sunny weather, they may not perform in inclement weather conditions. The real test for automated vehicle applications will be when the roads are wet or even icy and invisible to the sensors. Some of the challenges faced by the Automated Vehicles (AV) can be categorized into the following bins - Sensor Performance, Camera Performance, Vehicle Operational Parameters, Driver Behavior, Communications and Lack of Defined Standards. The AVs also rely on well-maintained data like general weather data, road weather data, crowd sourced data, and others like 3D maps. These data supplement the sensors and cameras in the AVs.

The United States Department of Transportation (USDOT) Road Weather Management Program (RWMP) has invested in several research efforts that support provisioning of road weather information in real time. The RWMP is developing several robust and comprehensive data management systems that are valuable and could be foundational in the research and development of AV applications for adverse weather conditions. The relevant data sets readily available from different data environments (e.g., Weather Data Environment) can be potential inputs to different AV applications. These inputs range from sensor data (air temperature, pavement temperature, etc.) to weather forecasts with temporal and geographic accuracy. These efforts can play an important role in the development of advanced safety applications for automated vehicles. In addition, the AVs could also serve as a source of road weather observations, given that AVs perform extensive sensing during their operations (e.g., RADAR, LiDAR, etc.).

This presentation discusses several aspects of this relationship between weather and Automated Vehicle (AV) performance in light of the active on-going work under the RWMP at the USDOT. The presentation will also provide a roadmap of the potential activities that could be considered to advance the symbiotic relationship between the RWMP and the AV research and development.