Now that we are in summer, the time par excellence of forest fires, any help is little to be able to abate them, especially in Spain, where on average around 28,000 hectares have been burned in the last decade each year. Or, what is the same, annually the equivalent of 40,000 football fields are burned in Spain. But to help mitigate the fires, a Portuguese company, Compta, and IBM Watson’s artificial intelligence (AI) have created Bee2FireDetection, a solution that allows to detect and fight fires in a much faster and more agile way. In addition, this tool can even calculate the probability of a fire occurring, before it does occur.
Bee2FireDetection uses artificial intelligence, spectrometers, different digital video surveillance cameras and thermography, a technique that allows remote temperature analysis to detect forest fires at distances of up to 15 km. 24 hours a day, 365 days a year. The collected images are analyzed with IBM Watson Visual Recognition using different deep learning algorithms, to allow the system to detect any sign of fire, such as a smoke column, an increase in temperature or a change in light, and give the alert practically in real time.
Calculation of fire risk
This pioneering system can also calculate the level of fire risk very effectively, combining Watson with a series of meteorological and environmental data from the terrain, such as relative humidity, wind direction or temperature. All this information is obtained from local weather stations and data from IBM’s The Weather Company, one of the most accurate weather forecasting platforms in the world.
By collecting and analyzing this vast amount of data, Bee2FireDetection can provide a very accurate forecast of the probability of a fire occurring, before it does. Knowing this risk index allows an adequate alert level to be established, and the fire fighting teams can be prepared to act in time. According to data from the Ministry of Agriculture, Fisheries and Food, in Spain this year almost 40,000 hectares have already been burned, a third more than the average since 2009. At the European level, June 2019 has been the hottest month of June since it was they have records. A situation that facilitates the appearance and spread of fires.
The Bee2FireDetection implementation does not require special hardware, it only needs an HD camera system and computers to monitor. IBM artificial intelligence also allows Bee2FireDetection to be a very useful tool in cases where a fire is already active. Taking advantage of Watson’s capabilities and the analysis of different meteorological and environmental data, the Portuguese company’s solution can estimate how the fire will evolve on the ground. The system can calculate, for example, the speed and direction the flames could take. Having this valuable and detailed information can help firefighters and other civil protection bodies to plan fire control in a very agile and efficient way.
According to data from the Ministry of Agriculture, Fisheries and Food, in Spain between January 1 and June 16, there have been a total of 6,220 forest fires, almost 4,000 more than in 2018. The arrival of this new solution can help to face this situation, and be especially useful in those communities where the fire alert is usually higher, such as Andalusia, Galicia, Catalonia, Asturias, Extremadura or Castilla la Mancha. Bee2FireDetection also has an industrial version, designed to detect and fight fires that can affect facilities such as factories, data centers and other important assets.
Bee2FireDetection joins other initiatives to improve fire fighting that IBM is involved in. An example of this is the Code and Response project, through which the American multinational promotes the development of open technology solutions that can help solve problems in those communities where help is most needed. Also, earlier this year IBM announced a solution to help companies in the energy sector to reduce the risk of fire by detecting vegetation that, due to its location or condition, can pose a security risk and cause a fire. Foliage, for example, can start a fire with some ease. Through a large amount of geographic and meteorological data obtained through satellites, drones and IoT sensors, each company can detect if there is any vegetation that can pose a security problem, and act accordingly.
