Flox Intelligence is developing a system based on edge artificial intelligence to monitor animals near railways, airports, roads, and critical facilities, then deter them with sound responses that vary according to the species and the animal’s behavior. The company says that around 150 of its devices are currently operating in North America, Europe, and Australia.
Flox spun out of an artificial intelligence laboratory at KTH Royal Institute of Technology in Stockholm in 2020 and was founded by Sara Nozkova after 13 years of work in transportation and autonomous vehicles. This year, the company is participating in the Startup Battlefield 200 finals as part of TechCrunch Disrupt, held in San Francisco from October 13 to 15.
How does the system work?
The battery-powered units can be mounted on poles, trees, or other structures. When a motion sensor detects nearby activity, it activates a camera that uses models running on the device itself to identify the animal. The device then emits a sound appropriate to the species in order to drive it away from the sensitive area.
Nozkova says the system changes its responses over time so that animals do not become accustomed to repeated signals. Flox also provides a software platform that displays device locations and what is happening in the field, with the ability to create reports for regulatory compliance and monitor biodiversity.
Why does this use case matter?
The need for these systems is based on the cost of animal collisions with infrastructure. The U.S. Federal Aviation Administration estimates that wildlife strikes cost civil aviation in the United States around $272 million annually on average between 1990 and 2025. In France, SNCF says that animals affect more than 8,000 trains annually, while the Brittany region saw 120 animal collisions with trains during 2023, resulting in the disruption of 660 services and more than 15,000 minutes of delays.
Flox’s approach differs from methods such as whistles, lights, fences, falcons, or removal, as it attempts to link the response to the animal’s identity and behavior while performing the analysis locally instead of sending monitoring data to the cloud. This may be important at remote sites or those requiring an immediate response, but the source does not provide independent measurements proving the system’s superiority over alternatives or the extent of its success in reducing collisions.
Expansion and competition
The company uses a business model that sells the units upfront and then charges an annual subscription for software and artificial intelligence after the first year. It also has OEM partnerships that enable its software to run on vehicles such as trains, including a partnership with Alstom. The platform is intended in the future to operate on trains, cars, drones, and other machines working outside controlled environments.
Flox faces competition from Robin Radar, which uses species-specific distress calls to disperse birds and record activity, and from iCHASE, which detects birds using computer vision and targets them with a moving laser. Nozkova says its advantage lies in the system’s breadth and the field data it has collected on safe responses for animals such as moose and grizzly bears.
Flox has raised around $5 million through pre-seed and seed rounds, led by Unconventional Ventures and Tilia Impact Ventures in the seed round, with participation from GENIUS NY, Norrsken, Satgana, and angel investors. The company is currently opening a $10 million Series A round, which will be directed primarily toward increasing the number of devices from around 150 to several thousand and expanding the system to uses beyond wildlife monitoring. The company expects to end 2026 with revenue of approximately $1.5 million.