Exploring preconditions for an integrated safe and smart traffic environment system - for Wildlife collision avoidance.

Norway has a large number of areas with wildlife danger along roads where it is not cost-effective to build large fences or wildlife crossings. The presence of wildlife along open roads can cause traffic collisions with wildlife, create stress and create dangerous situations for the safety of both people and animals. The seriousness of the situation is visible in the number of collisions with deer game. Despite attempts at various solutions, these have had little success. In total, more than 73,000 wild deer were hit on roads and railways from 2018 to 2022, and the number increases every year, figures from the Hjorteviltregisteret show. WILDETECT aims to collect both existing and new data to build knowledge about where the probability of collision accidents is highest, to warn road users and reduce the number of traffic accidents involving wildlife. The project will explore various solutions that collect data on the number and/or movement of ungulate groups, such as: 1) statistical data, 2) image observations with drones, and others 3) so-called unstructured data from road users. In addition, it will be explored how to bring data together (data fusion), to build knowledge about what type of data can be combined and how to improve statistical models to predict collisions. The knowledge generated in WILDETECT will also be used to test various warning systems, everything from mobile applications to variable speed limit signs that are activated when the risk is high. These warning systems will be evaluated by drivers. In order to test different warning solutions, WILDETECT will use a Virtual Reality (VR) based driving simulator to more effectively evaluate the drivers' behavior when the risk is warned. WILDETECT will also include actors involved in the transport system, including those who receive data, alerts drivers and road users. This will allow co-creation processes, where contributions from stakeholders and users will be considered to define the type of data that can be included in a solution. The project is currently collecting the knowledge status through literature review, interviews with relevant stakeholders, and early planning for workshops. NINA and HVL are currently preparing their data sources to be able to respectively predict collisions more accurately and to assess image quality of moose and deer. SINTEF is preparing the set-up VR-tests to be able to compare driving behaviour with different warning solutions and to prepare for real-traffic tests.

WILDETECT explores and extends various statistical and technical solutions in investigating the integration potential of data and technology in answer to traffic collision avoidance between vehicles and animals. The project aims to develop knowledge-based outcomes for the development of a safe and smart traffic system that can detect wildlife and warn road authorities and drivers of increased collision risk. The project results will be a starting point for developing future technology systems that can be applicable to traffic planning, providing road users with real and dynamic traffic information, thus contributing to increased transport safety.