Abstract Manager

Thousands of collisions with wildlife are reported annually on Canadian highways.  Through an analysis of historical collision information, Alberta Transportation and Economic Corridors identified a portion of Highway 2 as a corridor having a high density of Wildlife-Vehicle Collisions in the province.  To protect the surrounding natural wildlife habitat and improve safety for travelers several potential mitigation options were examined, ranging from civil infrastructure improvements to fencing.  The most appropriate solution was determined to be the application of sophisticated technologies to detect the presence of wildlife and advise motorists to slow down.

Accurately and reliably detecting wildlife on roadways and effectively actioning that data into useful notifications to drivers is a complex problem.  Factors such as weather, terrain, the unpredictable nature of wildlife, and the remoteness of these locations combine to make deploying an effective system very challenging.

The base solution consists of a suite of ITS components integrated with high-performance security technologies.  Sophisticated volumetric radar-based presence detectors monitor the 3-dimensional area of the roadway and the immediately adjacent shoulders to detect the presence of large animals on or in proximity to the roadway.  Thermal and color cameras record the presence of the animals and overall operation of the system to assist with calibration and verification.  A field-located central control system analyzes the detection data from all sensors.  When a large animal is detected, the control system activates a series of warning signs that advise motorists of the presence of animals and to slow down.

Most importantly, this project is being used as an opportunity to test multiple different Aritificial Intelligence (AI) detection systems in parallel with the proven radar-based system.  AI-equipped cameras can’t match the radar’s range, but can be deployed much quicker and for much lower cost on shorter sections of roadway.  If the AI technology can approach the detection effectiveness of the radar, this could revolutionize how animal detection is done, significantly improving rural road safety.  In addition to native AI detection capabilities on board the latest cameras that will be tested, PBX has undertaken extensive internal development in partnership with a leading AI provider to develop a cutting-edge transportation-specific AI-based animal detection application that meets the unique operational and data collection needs of transportation agencies.  The project is in construction and by the time of the conference configuration and system testing will be underway.  The presentation will include hot-off-the press observations, performance data, and lessons learned.