Wildlife strikes are an expensive and recurring problem for transport infrastructure. The FAA estimates they cost U.S. civil aviation about $272 million a year on average between 1990 and 2025; France’s SNCF says wandering animals affect more than 8,000 trains annually and rank among its five leading causes of delays. In Brittany alone, 120 trains hit animals in 2023, disrupting 660 services and causing more than 15,000 minutes of delay. Sara Nozkova, who saw the aftermath of a train hitting a wild bison on a childhood trip to Mumbai, concluded that traditional deterrents — sirens, lights, fences, falcons, and culling — were the wrong starting point. “I was convinced the answer was not a louder deterrent,” she told TechCrunch.

After 13 years in transportation and autonomous vehicles, Nozkova joined the AI lab at KTH Royal Institute of Technology in Stockholm to build a wildlife-deterrent system that changes its response based on what an animal is doing and even what species it is. That work became Flox Intelligence, spun out of KTH in 2020 and now a finalist at TechCrunch Disrupt’s Startup Battlefield 200. Flox makes edge devices that process data on-site rather than in the cloud. Like autonomous vehicles, they use perception systems, but those systems are trained on wildlife behavior. A motion sensor wakes the unit, a camera captures images, and on-device models identify the animal before emitting a species-appropriate sound. Nozkova says the system learns from experience and varies its responses so animals do not get habituated to repeated signals.

The current product line includes battery-operated devices that can be mounted on poles, trees, and other structures, plus a software platform showing where devices are deployed and what is happening on the ground. The platform can also generate reports for regulatory compliance and biodiversity monitoring. Roughly 150 Flox devices are live in North America, Europe, and Australia, at airfields and airport perimeters, road and rail corridors, national parks, agricultural land, urban areas, mining sites, and electrical substations. Flox sells edge units up front and charges an annual software and AI subscription after the first year. It also has OEM partnerships, including with train manufacturer Alstom, which pays an annual fee to deploy Flox software on vehicles. Nozkova says the company expects to end 2026 with revenue of about $1.5 million.

Flox has raised about $5 million across pre-seed and seed rounds, with the seed round led by Unconventional Ventures and Tilia Impact Ventures. Co-investors include GENIUS NY, Norrsken, Satgana, and angel investors. The company is opening a $10 million Series A, which would mostly go toward growing from about 150 devices to several thousand deployed units and extending the system beyond current wildlife use cases. The longer-term goal is making the software portable enough to run on machines that work outside, including trains, cars, and drones.

Competition exists: Robin Radar’s Scarecrow uses species-specific distress calls and records activity, while iCHASE detects birds with computer vision and targets them with a moving laser. Nozkova argues Flox’s moat is the breadth of the system and the field data it has collected, which she calls “ground truth on what actually makes a moose or a grizzly bear move safely.” Her broader pitch ties the product to a shift in physical AI: “Every machine that operates outside — trains, trucks, drones, tractors, robots — will need to understand nature, and nature intelligence is the next layer of the perception stack. Almost nobody is building it.”