Development of a Real-Time Roadway Debris Hazard Spotting Tool Using Connected Vehicle Data to Enhance Roadway Safety and System Efficiency
-
2019-08-31
Details:
-
Creators:
-
Corporate Creators:
-
Corporate Contributors:
-
Subject/TRT Terms:
-
Resource Type:
-
Geographical Coverage:
-
Edition:Final
-
Corporate Publisher:
-
Abstract:Using the information residing in connected vehicle Basic Safety Message (BSM) data, this project develops an algorithm that is capable of identifying the location of unwanted objects, i.e. debris on road segments. Vehicles’ lateral accelerations are used to detect lane changing and swerving behaviors to create density maps that pinpoint the locations with high frequency of swerving behavior and thus debris locations. Two vehicles were used to collect the required data on a selected US DOT Tampa CV Pilot road segment. A portion of the data was used to fine-tune the algorithm parameters and the rest was used to test its ability to locate the object on the road. The overall accuracy of the algorithm to detect individual lane changes is estimated to be 96 percent. Coupling the algorithm with density diagrams, debris locations could be identified almost precisely. The algorithm has the potential of reducing time and money spent by state and local agencies to swipe and remove debris from the road, as well as reducing the risk of crashes caused by drivers’ swerving behavior to avoid on-road debris.
-
Format:
-
Funding:
-
Collection(s):
-
Main Document Checksum:
-
Download URL:
-
File Type: