Flox Intelligence Deploys AI-Powered Wildlife Detection to Protect Infrastructure
A Stockholm-based startup is using edge AI and species-specific deterrents to prevent animal collisions with trains, planes, and vehicles—a problem that costs the aviation industry alone $272 million annually.
Wildlife collisions with transportation infrastructure represent a persistent global problem. The U.S. Federal Aviation Administration calculates that such incidents have averaged $272 million in annual costs to civil aviation between 1990 and 2025. France's national rail operator SNCF reports that animals wandering onto tracks disrupt more than 8,000 trains yearly and rank among the five primary causes of service delays. In the Brittany region alone, 120 animal strikes occurred in 2023, resulting in 660 service interruptions and exceeding 15,000 minutes of accumulated delays.
These statistics motivated Sara Nozkova to create an alternative to conventional wildlife management approaches. Rather than relying on sirens, lights, fencing, trained falcons, or culling, Nozkova developed a system that adapts its deterrent strategy based on animal behavior and species. "I was convinced the answer was not a louder deterrent," she told TechCrunch.
Nozkova's background in transportation and autonomous vehicles shaped her approach. Over 13 years in those fields, she observed how machines trained for controlled environments struggle when confronted with natural complexity. "We have taught machines to be brilliant, but only inside the environments we engineered for them," she said. "Cities, factory lines, warehouses. Fixed geometry, known objects. Stepping outside into nature, I saw firsthand how the same machines quickly go blind. Because nature is complex and always alive."
This insight led her to collaborate with the AI laboratory at KTH Royal Institute of Technology in Stockholm on a system capable of adjusting deterrent responses according to animal activity and species identification. The resulting company, Flox Intelligence, spun out from KTH in 2020 and now competes in the Startup Battlefield 200 at TechCrunch Disrupt, scheduled for October 13-15 at Moscone West in San Francisco.
How Flox's Technology Works
Flox manufactures edge devices that process information locally rather than transmitting to cloud servers. Similar to autonomous vehicles, these units employ perception systems trained specifically on wildlife behavior. When a motion sensor detects movement, it activates a camera equipped with on-device models to classify nearby animals. The device then broadcasts a sound calibrated for that particular species to frighten it away.
https://www.youtube.com/embed/SqkhdhuN6OU?si=NJXmlR-AkIzvllDM&autoplay=1&mute=1
The system incorporates learning capabilities and modulates its responses to prevent animals from becoming accustomed to repeated signals. Battery-powered units mount on poles, trees, and other structures. Flox also provides a software platform displaying device locations and real-time ground conditions. This platform generates reports supporting regulatory requirements and activities such as biodiversity assessment. Future iterations aim to make the software compatible with mobile infrastructure including trains, vehicles, and drones.
Approximately 150 Flox devices currently operate across North America, Europe, and Australia. Deployment locations include airfield perimeters, road and rail corridors, national parks, agricultural zones, urban environments, mining operations, and critical infrastructure like electrical substations.
Business Model and Revenue
Flox sells edge units upfront and charges annual subscription fees for software and AI services beginning in year two. The company maintains original equipment manufacturer partnerships with organizations that deploy its software on vehicles; train manufacturer Alstom represents one such partner, paying annual licensing fees. Nozkova projects the company will achieve approximately $1.5 million in revenue by the end of 2026.
Competitive Landscape
Flox faces competition from established players in wildlife deterrence. Robin Radar's Scarecrow system employs species-specific distress calls to disperse birds while recording activity. iCHASE offers a computer vision-based detection system that targets birds with an adjustable moving laser.
Nozkova contends that Flox's competitive advantage stems from its comprehensive system architecture and proprietary field data. She describes this accumulated knowledge as "ground truth on what actually makes a moose or a grizzly bear move safely."
"Physical AI is walking out of the factory and into the messy, alive outdoors," Nozkova says. "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."
Funding and Growth Plans
Flox has raised approximately $5 million through pre-seed and seed funding rounds to date. Unconventional Ventures and Tilia Impact Ventures led the seed round, with additional backing from GENIUS NY, Norrsken, Satgana, and individual angel investors.
The company is now launching a $10 million Series A funding round. Proceeds will primarily support scaling from the current 150 deployed devices to several thousand units and expanding the system's applications beyond wildlife management.