Around 330,000 people have been forced to flee their homes and holiday destinations as some of Europe’s worst wildfires of the summer spread across the Iberian Peninsula and southwestern France, destroying villages and incinerating hundreds of thousands of acres of forest. In France’s Gironde region near Bordeaux and in wildfire-hit areas of Spain, flames are being fueled by extreme heat — and scientists say some fires are now creating their own weather systems, making them even harder to control.
By the time flames are visible on the horizon, it is often already too late to stop them. That reality — the gap between ignition and detection — is exactly where Hong Kong-based Robotics Cats comes in. Its AI vision software technology turns ordinary cameras into an early-warning system designed to spot a wildfire’s first wisp of smoke before it becomes headline news.
Robotics Cats core offering LookOut Wildfire Detection software-as-a-service (SaaS), uses computer vision and deep learning to detect early-stage wildfires from smoke and flame signatures, drawing on a training set of more than 500,000 wildfire images. The company also offers InsightFD, a scanning system aimed at government. The company claims its algorithm can identify a wildfire signature as small as 0.03% from a source image and detect smoke or fire at distances of up to roughly 15 kilometers.
Customer segments include national parks, forestry and wildfire agencies, UNESCO World Heritage sites, private plantations, power/telecom infrastructure providers and tourist resorts, as well as individual homeowners.
“AI can do much better than humans at monitoring wildfires and if AI finds the fire early, we can control them much more successfully,” says CEO and founder Andre Cheung, who previously served as business development director at Insight Robotics, where he sold the InsightFD wildfire detection system.” Prior to that Cheung, who holds a computer engineering degree from the University of Hong Kong, spent more than two decades in the telecommunications and cloud computing sectors in the Asia Pacific region.
The seven-year-old Hong Kong-based startup, which currently operates in 13 countries in Asia, Europe and the Americas, was named a top innovator by the World Economic Forum’s UpLink platform at the Forum’s Annual Meeting of the New Champions in 2024. Separately the company is part of a four-company coalition which qualified in the final round of the $11 million XPRIZE Wildfire competition by combining satellite, ground-camera, Internet of Thing (IoT) and drone suppression technologies. (The other members of the coalition ares Indicium Dynamics of Australia, Little Place Labs in the U.S. and Australia’s Taz Drone Solutions)
Traxn lists nearly 70 companies in the forest wildfire detection category. These companies broadly fall into three camps: ground-based camera/computer vision networks such as Robotics Cats and a U.S. competitor called Pano AI; satellite infrared monitoring, such as Germany’s OroraTech, and companies such as Berlin-based Dryad Networks, which. makes distributed IoT sensor meshes that detect gas and heat signatures at the point of ignition.
One of Robotics Cats key points of differentiation is that it needs no new hardware. While competitor Pano AI deploys its own proprietary 360-degree units, Robotics Cats core pitch is that it works on cameras a customer already owns. It is a technology integrator partner of Axis Communications, a Swedish company specialized in IP-based security cameras installed in airports, shopping malls, highways and stadiums but the Hong Long company’s technology also works on virtually any high resolution camera, Cheung says.
The company, which has so far raised $600,000 from Beyond Ventures and Hong Kong’s Innovation and Technology Fund, additionally has an adjacent product line that applies the same computer vision stack to count and identify endangered wildlife species, positioning the company beyond pure fire detection. It is also developing AI technologies for early warning of other disasters , such as typhoons and extreme heat, which wreak havoc and cause increasing damage across Asia.
Robotics Cats is already collaborating with Pavlou Fire Fighting Co. in Greece to deploy its technologies and camera systems in Athens to fight wildfires.
Asked why AI has not been more successful in slowing this summer’s wildfires in Europe Cheung said “government procurement processes are relatively slow, it takes time for them to act,” he says. The private sector is increasingly stepping in: “Energy and infrastructure firms are organizing and participating in innovation programs to speed up technology adoption and connect startups with government,” says Cheung. “I’m optimistic that FireTech adoption will increase in Europe.”
