Detect natural disasters and coordinate rapid response actions is the purpose of a computer system developed by scientists of the Massachusetts Institute of Technology and the Open University of Catalonia.

This system has been trained with a database of over 1.7 million photosto be able to afterwards automatically detect natural disasters through images posted on social networks.

This is “Campi Flegrei”, the supervolcano on the brink of its first eruption since 1538, which could spark a possible “worldwide winter”

“We showed that automatic incident detection on networks such as Twitter is possible, and this can greatly help humanitarian aid organizations”, emphasized the UOC professor Ágata Lapedriza, who specializes in artificial intelligence. Since the consequences of natural disasters cannot yet be predicted, the expert says It is important to articulate a rapid and effective emergency response and international cooperation to save lives.

Some of the events that can be devastating are floods, tornadoes or wildfires, which, on the other hand, are becoming more and more frequent and devastating due to climate change.

“Fortunately, technology can play a very important role in these situations. Social media posts can be used as a low-latency data source to understand the progression and impact of a disaster”explained Lapedriza.

Where does the El Niño phenomenon occur and how does it arise?

A total of 43 categories of natural incidents (avalanches, sandstorms, earthquakes, volcanic eruptions, droughts…) and different types of accidents with some degree of human intervention (aircraft, construction works, etc.) have been created for the system. ). Each image has a label that helps you distinguish a bonfire from a fire, for example.

“Within these images, our model detected the images that matched incidents and matched specific incidents for which there was a record, such as the 2015 Nepal and Chile earthquakes,” explains the professor and researcher.

“This opens doors for humanitarian aid organizations to more efficiently find out what is happening and to improve the management of humanitarian aid where necessary.”, added the specialist.

The study is published in the journal Transactions on Pattern Analysis and Machine Intelligence.