How “smart” technologies help in the fight against forest fires – ΥΣΙΑ

The director of MTS in Yakutia, Prokopy Neustroev spoke about the existing effective digital solutions

In world and Russian practice, there are already effective digital solutions for fighting and preventing fires. Exactly how, told us the director of MTS in Yakutia, Prokopiy Neustroev.

The forest fires in the country in 2022 marked an anti-record. There are currently 18 forest fires in Yakutia. The total area of ​​all fires covered by fire is almost 13 thousand hectares. Traditional disaster prevention and mitigation technologies often prove ineffective: they are not always accurate, they can not process large amounts of data quickly, etc.

According to the director of MTS in Yakutia Prokopios NeustroevaIn this regard, forest fire risk management is an urgent task not only for specialized organizations and companies, but also for computer programmers who create alternatives, often based on artificial intelligence, which help in the detailed study of large volumes of geospatial, meteorological and temporal data for the prediction of the occurrence of fires.

Global experience

South Korea. Companion against fire and criminals

In 2019, an employee of Hongik University in South Korea (Jaik Seung Lee) and his students developed an algorithm that can predict the occurrence and behavior of fires with 90% accuracy. The technology is based on high-resolution satellite imagery and Azure Machine Learning algorithms from Microsoft.

The creators of the system claim that with its help, firefighters can more rationally allocate resources: to reorganize the brigades, to send them where there are not enough resources. This allows services to respond more quickly to calls and minimize potential fire damage. Jae Seung Lee hopes that this tool can be used for other purposes, such as the fight against crime.

China. No man is an island

A joint forest fire prevention project was developed and implemented by Huawei and Enbo. To determine the exact location of the smoke and prevent fires in advance, the cameras analyze the received videos, the main analysis is performed by Atlas products and the secondary analysis by cloud services.

By analyzing data from video cameras, it is possible to make accurate predictions for the occurrence of forest fires and to determine their location in time. The system is fully automated and does not require the participation of a large number of people – for example, in Guangdong Province, where it already operates, only one person monitors and manages nearly 1000 camcorders. The patrol cycle lasts from 3 to 10 minutes and the fire is detected on average 2 hours earlier than usual.

USA. Drones and medical developments

In 2019, CrowdAI introduced the FireNet program, which uses artificial intelligence to predict forest fires. The algorithm analyzes images from digital drones that circle over trees and take pictures at 20 frames per second. The accuracy of fire detection is 92%. Data processing is done in real time. The points with a high probability of fire are determined by the GPS coordinates of the drone.

According to Vice, when FireNet was created, the experts were inspired by medical image analysis programs that help doctors distinguish healthy tissue from unhealthy tissue. The system was trained in video recordings of forest fires taken by unmanned aerial vehicles.


Smart cameras and heat sensors

Every year we hear about forest fires in the Far East and Siberia. The introduction of intelligent systems in the field of natural area protection can serve as a serious helper to the regions. Such solutions already exist, for example, at MTS, which is actively involved in developments based on artificial intelligence.

In Russia, in the Sebezhsky National Park in the Pskov region, operates a smart video surveillance complex, which monitors the state of fire hazard and was introduced by the MTS. The complex consists of fourteen cameras with special lenses and sensors. Webcams are monitoring forest fires, recording a dangerous rise in temperature that allows state park inspectors to prevent wildfires at an early stage, even before an open fire breaks out. The thermal imagery also informs the monitoring center of fire and smoking events in the protected area.

The MTS set of innovative digital systems helps ensure 24-hour surveillance and safety in the park: early detection and fire prevention, as well as assistance in finding people and creating comfortable conditions for visitors to the national park.

“The company has solutions based on computer vision technologies and colleagues from other areas have successfully implemented several cases in specially protected natural areas. They rely on artificial intelligence, with the help of which machines and computers can see and understand what is happening in the real world. “In Yakutia, we can also introduce firefighting technologies and other ‘smart’ developments that have already proven effective,” said Prokopiy Neustroev.

Leave a Comment