As one of the most unpredictable phenomena in nature, lightning is very disruptive. Scientists have recently developed an AI system that predicts lightning up to 30 minutes before it strikes.
Lightning regularly kills people and animals, starts fires, damages power lines and keeps aircraft grounded. Until now it has been virtually impossible to predict lightning, with no simple technology for predicting when and where it will strike the earth.
Engineers at the Ecole Polytechnique Federale de Lausanne’s (EPFL) School of Engineering developed a simple and inexpensive system to predict when lightning will strike. The research led by Farhad Rachidi, resulted in a method of predicting lightning between 10 and 30 minutes before it strikes, within a 30km radius.
Using a combination of Artificial Intelligence and meteorological data, researchers are now planning to use this technology in the European Laser Lightning Rod project, a project designed to draw lightning away from areas that are susceptible to lightning damage, the project is shown in the video bellow.
“Current systems are slow and very complex, and they require expensive external data acquired by radar or satellite,” explains Amirhossein Mostajabi, the PhD student who came up with the technique. “Our method uses data that can be obtained from any weather station. That means we can cover remote regions that are out of radar and satellite range and where communication networks are unavailable.”
Due to the ability to acquire the data in real time, the method allows meteorologists to quickly predict lightning before a storm forms and alert those who could be affected.
The method created by EPFL researchers uses a machine learning algorithm that has been trained to detect the conditions that lead to lightning. Researchers considered four parameters when creating their method; atmospheric pressure, air temperature, relative humidity and wind speed. These parameters are correlated with recordings from lightning detection and location systems.
After the system has been trained, researchers believe the method to be 80% accurate when predicting the location of a lightning strike.