Developing a situation and threat assessment framework for a next generation roadside animal detection system

Abstract

Collisions involving large animals are a serious safety, economic and ecological concern. Some North American jurisdictions have installed a roadside animal detection system (RADS) that can warn the possible presence of large animals on rural highway sections. This study provides a conceptual framework for developing a next generation (NG) RADS. This study focuses on developing a process that can estimate the varying levels of threat posed by animals on the roadway using real-time data on animal and vehicle positions. To estimate the level of threat, the study used a fuzzy rule-based algorithm that integrates four input indicators (e.g., physical distance between animal and vehicle). The methodology was tested using real-world traffic and animal data collected from a conventional RADS in British Columbia, Canada. The NG RADS has significant advantages over the conventional RADS. In particular, the NG RADS can disseminate varying levels of warning according to the estimated level of the threat rather than the constant level of warning generated by a conventional RADS. The NG RADS can also use a Vehicle-to-Infrastructure communication technology to establish direct wireless communication with vehicles at risk, for instance, to automatically control a vehicle's speed to avoid a collision with a large animal.