Google is working on a project that aims to make use of artificial intelligence (AI) technology in predicting the locations of earthquake aftershocks. The technology is still premature at present, but it could be a big help to many people in the future. 

A scientist from Google Inc. disclosed Tuesday that Google is developing a technology that utilizes machine learning to make forecasts of earthquake aftershocks. The company started the project after realizing the need to locate aftershocks so that recovery efforts could be carried out efficiently following the mainshock. 

“We applied machine learning algorithms to analyze a database of earthquakes from around the world to try to predict where aftershocks might occur,” Martin Wattenberg, a senior staff research scientist at Google’s People + AI Research (PAIR) initiative, was quoted as saying by Yonhap.

Wattenberg is currently working together with Harvard University researchers in developing the technology that could predict where aftershocks might take place. According to the scientist, the system they came up with so far has a precision rate of six percent, which is better from the previous three percent. 

Korea Herald has learned that the neural network system used by the researchers was trained on over 131,000 earthquake data. Interestingly, it’s already capable of predicating aftershock locations in an independent test data set of over 30,000 earthquakes more accurately than the Coulomb method. 

“It’s still a low precision … it’s not accurate enough for practical use but it’s an exciting first step,” Wattenberg said. “This is a real improvement, but to be clear, not yet strong enough for practical use.”

While there’s still a long way to go before the technology becomes quite accurate in locating aftershocks, the Google scientist is optimistic that it could one day become useful in emergency and rescue operations, as well as in evacuation plans. 

Google Google is developing an AI technology that predicts earthquake aftershocks’ locations. Photo: Getty Images/Johannes Eisele