Gubkin University Develops AI System to Respond to Fires at Energy Facilities
It can respond within seconds.
Researchers at Gubkin University have developed an AI system designed to improve safety at fuel and energy facilities. The system can cut response times to fires by a factor of 10.
According to the developers, deploying the technology could reduce fire-related damage by at least 58%. Previously, neural networks used in industrial safety could only determine whether a fire was present. The new system takes a fundamentally different approach: instead of simply detecting a fire, it performs an analysis of what is happening.
The system processes images from surveillance cameras and assigns each incident a hazard level. The team tested it on images of real fires recorded at an operating fuel and energy facility. As a result, the operator’s response time fell from one to two minutes to 10 seconds.
Importantly, the algorithm filters out false fire-alarm triggers. This significantly reduces the workload for operators. They no longer need to review huge amounts of information because the system generates a detailed, structured report on the incident.
