
Source: BFM-Business

Early discovery makes it possible to preserve a building from a fire, a forest from a devastating wildfire, or to secure industries.
Iris Fire, a subsidiary of Iris Tech, operates in France and Europe to conduct research and technological foresight studies. Video fire detection systems with artificial intelligence. The discussion with David Belin, founder of Iris Tech.
I am an IT network engineer. Iris Tech is a company specializing in the installation, security, and management of IT infrastructure and networks for small and medium-sized businesses (SMEs). This is my core business. For the last ten years, we have been developing the innovative fire detection system with our subsidiary Iris Fire, because there is a real demand for innovative technologies in this sector. We work with various companies, both German and American, to adapt the cameras so that they match the internal architecture.
We provide various algorithms based on external, internal, environmental and risk types, as well as distance. We assist our clients through the various stages of their project: architectural and needs analysis, physical installation of the device, software configuration and maintenance.
Iris Fire specializes in integrating innovative video analytics systems coupled with AI for the detection of fire, smoke, and oil mist. We work with various technologies for fire detection.
For example, we offer an external camera, specific for forest fires, that is mounted on high towers. Another one used in the navy allows the identification of oil mist and smoke. A triple AI camera will be used in the landfills for the early detection of fires. For the nuclear field, we have cameras that are resistant to gamma rays (red zone).
Regarding monitoring, the hybrid system incorporates the latest technologies allowing for optimal surveillance (raising doubts, precise geolocation of incidents, export of videos..), while also having the ability to interface with more conventional systems (conventional fire control center, emergency response team, etc.). We adapt to the demand.
For each problem, we have different technologies, algorithms, or sensors. For example, for forest fire detection, we have formed a partnership with a German company that adapted the technology from the Rosetta mission for ground-based day and night detection. The 360-degree camera uses 4 sensors. It covers 120,000 hectares, a radius of about 40 kilometers. This technology allows for early detection of fires and, thanks to triangulation, for pinpointing the fire location. GPS coordinates are sent directly to the SDIS for an early intervention.
A small leak on a boat engine causes the appearance of micro droplets of oil that, when mixed with air, can cause an explosion. Our cameras are placed sufficiently far, in ATEX enclosures, and offer the possibility of detecting oil mist early enough to prevent an explosion. The same applies to waste disposal sites where methane and ammonia can cause a fire. The triple IA algorithm detects a spark; it is extremely accurate, unlike thermal analysis, which only provides a detection of a change in the temperature threshold. However, when a thermal vehicle passes nearby, the temperature threshold increases, which generates many false positives. The video analysis processes the pixel change using standardized models stored in a virtual neural network. For example, we have equipped the La Fayette stealth frigates of the National Navy with oil mist detection.
We also installed a system in Saint-Avold for Total-Arkema to protect the power plant inside the factory. With conventional detection sensors, they would have been forced to completely stop the electrical production. However, in this area, electric arcs are capable of melting IPNs. We were able to install our cameras outside the area, for protection inside the area without entering it. .
We are the first to have implemented this type of fire detection in France, so we will continue to advance technological innovation and expand these innovative technologies to become a benchmark at the European level.
This content was produced in partnership with SCP. The BFM Business editorial team did not participate in the production of this content.