The possibilities of the use of Artificial Intelligence in fire and hurricane forecasting and personalized agriculture, presented by the professor of the Sloan School of Business Administration of the Massachusetts Institute of Technology, Dimitrios Bertsimas, during his speech at the conference on “Climate Resilience and Local Society”, held in Alexandroupolis, at the initiative of the President of the Republic, Katerinas Sakellaropoulou, and organized by the Eugenidou Foundation.

Originally from Alexandroupoli, Mr. Bertsimas watched with great interest and concern, for the fate of his relatives, the disastrous fires the previous summer in Evros.

He and his scientific team, consisting of about 40 people, including Greeks, worked on the application of Artificial Intelligence to fires in California, USA.

With data on weather, humidity, wind, precipitation and other information, they built a system that decides on the optimal placement of ground sensors to maximize coverage while also taking into account their budget.

They built algorithms and a wildfire risk model.

Mr. Bertsimas stated that he is ready to make available the algorithms and this entire Artificial Intelligence application system to be used in Evros and in Greece in general, stressing, however, that the most important thing is the coordination of the people and the services that will use it of these data.

“In the case of fires we have data from satellites, we have data from sensors on the ground, we have atmospheric data and we also have drones that we can manage to get information in a continuous way.

When one uses all this data together one can apply the idea of ​​Artificial Intelligence, make real-time inferences about wildfires, hurricanes, agriculture and many other applications.

Artificial Intelligence can find solutions to the early detection of fires by improving early detection systems that monitor and alert authorities to emerging fires in real time.

By collecting data from satellites, drones, heat sensors, smoke sensors, it is possible to predict the progress of fires in real time. And then lead to decisions about where to put the sensors, how to guide the drones.

Early detection of fires can lead to early mobilization of authorities, allocation of resources, evacuation planning, protection of firefighters, communities, ecosystems and the environment,” Mr. Bertsimas pointed out.