Radar-Based Animal Detection System on U.S. Highway-95 Reduced Traffic Speeds by 0.7 to 4.4 Miles per Hour.

Researchers Test an Animal Detection System in Bonners Ferry, Idaho.

Date Posted

The Reliability and Effectiveness of a Radar-Based Animal Detection System.

Summary Information

This study evaluates the reliability and effectiveness of a doppler radar-based animal detection system. Reliability was measured by how well large mammals (deer size and larger) were detected. System effectiveness was measured by reductions in vehicular travel speed; speed radars were also installed along the road section to assess whether drivers had sufficient time to respond to an activated warning sign. The study took place from Fall 2015 to Summer 2016.


To assess system reliability, thermal video cameras were used to monitor wildlife in the detection area, a 371 ft long, 72 to 121 ft wide road section of U.S. Highway 95. When the radar detected a “large mammal,” the system took a thermal image every three seconds. An amber light placed above warning signs north and south of the detection area would be activated for 40 seconds. Drivers were also informed of the detection zone with signs displaying “WILDLIFE DETECTION TEST AHEAD.” A walk-through test and seasonal reliability tests were conducted in 2016 to assess reliability. In addition, three speed radars were installed along the road section to record vehicle speed to determine whether warning signs provided enough time for drivers to respond to the activated sign. Experimental design based on treatment-control groups was used to compare vehicle speed with the warning signs off versus on.


  • Vehicle speeds decreased 0.7 to 4.4 miles per hour (mi/h) while warning signs were activated in autumn and winter.
  • The radar-based system correctly detected 76 percent of animal crossing events.
  • The false negative rate was estimated at 2.5 percent.
  • For 75 percent of the deer, the warning signs were “on” for the entire time the deer was on the pavement, and for 90 percent of the deer the warning signs were “on” the entire time or for part of the time the deer was on the pavement.
  • 100 percent of elk were detected and all of the warning signs were activated while elk were on the pavement.
  • Up to 68 and 85 percent of deer were detected early enough for most northbound drivers (58 to 68 percent) and most southbound drivers (70 to 85 percent) to respond to the warning.

Although a before-and-after crash analysis was performed, the evaluation of safety benefits was limited due to the short length of the road section equipped with the system. functionality, which costs an additional $3,000 per 250 meters annually.

Goal Areas
Results Type