Railway Technology Artificial Intelligence Reduces Train Collisions with Wildlife

Related Vendors

dataTec AG

Taiwan Machine Tool & Accessory Builders' Association (TMBA)

SIMTOS (Seoul International Machine Tool Show)

Alstom and Flox Intelligence are currently testing in Sweden how artificial intelligence can detect and deter wildlife near railway tracks. The initial results are positive —the system will help reduce the number of accidents.

Wildlife collisions are one of the most common causes of accidents in rail traffic in Sweden, with around 5,000 collisions with wild animals occurring each year. To reduce this number, the European rail technology provider Alstom and the Swedish deep-tech startup Flox Intelligence are conducting a field test in Sweden for the first time with an AI-powered system that detects animals in real time and deters them with acoustic signals if necessary. According to Alstom and Flox, this approach is intended to make rail operations safer and more punctual, as well as improve the working conditions of train drivers. Additionally, the new development aims to reduce repair costs for trains and preserve biodiversity along the tracks.

The system is based on AI-powered, real-time cameras. As soon as they detect a collision risk, they trigger specially tailored audio signals to scare animals away from the tracks. In an initial project phase, the artificial intelligence identified a number of animal species, such as moose, deer, wild boars, and foxes. In April 2026, the second implementation phase began, during which Alstom and Flox are testing the complete system of video cameras and acoustic deterrence.

"Our field tests show that AI has the potential to identify animals with high precision. By combining our expertise in wildlife deterrence with Alstom's leading role in rail innovation, we are creating a solution that protects animals while also increasing operational safety," emphasizes Sara Nozkova, CEO of Flox Intelligence.

Unexpected Insights into Wildlife along the Tracks

The system uses artificial intelligence, bioacoustics, and behavioral science to detect and drive wildlife away from the tracks. Each animal detection is categorized to improve future identifications. During the trials, the detection of livestock and birds such as crows or pigeons proved particularly accurate. However, moose and deer required more intensive training to achieve the same level of detection accuracy. Both large animals and smaller species and birds, which were rarely statistically recorded before, were also detected. "The tests so far have given us a better understanding of which animal species are present near the tracks, how effective the existing wildlife fences are, and how the technology can contribute to both safer traffic and new insights into wildlife along the railway," explains Maria Signal Martebo, Managing Director of Alstom Sweden.

Tests on Several Swedish Railway Lines

The test runs are being conducted together with the regional railway company Tåg i Bergslagen (TiB) and its operator VR, a subsidiary of the Finnish state railway, on various routes. These include the Dalabanan, the Bergslagsbanan, the Godsstråket through Bergslagen, and the Bergslagenpendeln.

The project is funded by the Swedish innovation agency Vinnova with 3.3 million Swedish kronor, which corresponds to approximately 300,000 euros. It is considered an important step in Sweden's and the European Union's efforts to make railway systems safer and more sustainable.

The Innovation Station of Alstom and Tåg i Bergslagen already signed a partnership agreement in April 2024. In addition to the now-tested wildlife detection, this agreement also includes the use of operational and condition data with the help of AI, the testing of new technologies for passenger information systems, and the application of additive manufacturing using 3D printing.

Additive manufacturing

Basic knowledge 3D printing