Evaluating Image-Derived Estimates of Road Weather Conditions

Authors: Brittany Welch*, , John Horel, University of Utah
Topics: Transportation Geography, Hazards and Vulnerability
Keywords: transportation, road weather, image processing
Session Type: Paper
Day: 4/9/2020
Start / End Time: 9:35 AM / 10:50 AM
Room: Capitol, Sheraton, IM Pei Tower, Terrace Level
Presentation File: No File Uploaded


Winter weather creates hazardous driving conditions nationwide that lead to road weather fatalities and crashes. State transportation departments (DOTs) have to balance winter roadway safety relative to their available resources. For example, the Utah DOT spends 25 million dollars annually for winter maintenance operations. Road Weather Information Systems (RWIS) are one automated approach to monitor weather and pavement conditions continuously. Although there are approximately 150 RWIS across Utah and Wyoming, those observing sites are insufficient by themselves to monitor the conditions along the extensive roadways in those states.

DOTs have invested heavily in cameras installed along roadways to provide weather, road condition, and other factors affecting traffic flow. There are so many cameras in operation within each state that it can be difficult for staff to be aware of adverse conditions at all times at all places. Surface transportation products, such as the HeliosĀ® Real-time Ground Weather Intelligence System of the Harris Corporation, are available to monitor weather and pavement conditions obtained from camera networks. Helios relies on image processing to provide transportation staff with web services to monitor road weather information throughout their area of responsibility.

The accuracy of the camera-based estimates of pavement state, visibility, and precipitation available from Helios is being compared to observations of those parameters at nearby RWIS sites during the 2018-2019 and 2019-2020 winters in Utah and Wyoming. The ability of the Helios output to correctly identify road surfaces that are fully or partially covered by snow is examined.

Abstract Information

This abstract is already part of a session. View the session here.

To access contact information login