We all are aware how catastrophic earthquakes can be and how catastrophic the after maths can be. For the longest time we didn't have a technology that could tell the areas that will be affected in the earthquake aftermath.
Now an ingenious Artificial Intelligence application has been developed by the scientists that will be able to predict the place which will be impacted by post earthquake vibrations. Harvard University is leading a team of researchers that will work on an AI that is trained to crunch large quantity of sensor data and apply deep learning to make more accurate predictions. According to the researchers working on the technology "the system is not ready to be deployed yet, but is already more reliable at pinpointing aftershocks than current predictions models." Brendan Meade from Harvard University says "there are three things you want to know about earthquakes you want to know when they are going to occur, how big theyâ??re going to be and where they are going to be." He further added "Prior to this work we had empirical laws for when they would occur and how big they were going to be and now we're working the third leg, where they might occur."
Why this technology is perfect for earthquakes?
With so many variables to consider from the strength of the shock to the position of the tectonic plates to the type of ground involved. Deep learning could potentially tease out patterns that human analysts could never spot. To put this to use with aftershocks, Meade and his colleagues tapped into a database of over 131,000 pairs of earthquake and aftershock readings, taken from 199 previous earthquakes. With the help of AI they then were able to predict the activity of more than 30,000 similar pairs, suggesting the likelihood of aftershocks hitting locations based on a grid of 5 sq. kms. units. The new AI system managed to beat the current system "Coulomb failure stress change". While Coulomb model scored an accuracy of 0.583 which close to a flip coin i.e. it might happen it might not happen, the new AI system managed to score 0.849 which is the closest it any thing has come to perfect accuracy of 1.
Phoebe DeVries one of the researchers from Harvard University says "I'm very excited for the potential for machine learning going forward with these kinds of problems it's a very important problem to go after. Aftershock forecasting in particular is a challenge that's well-suited to machine learning because there are so many physical phenomena that could influence aftershock behaviour and machine learning is extremely good at teasing out those relationships."
Why it is not being put into use yet?
There are a few reasons because of which this system is not being used, one of the researchers pointed out by saying that the current model is only designed to deal with only one type of aftershock trigger and simple fault lines, it is yet not ready to be applied around the world. Another reason of yet not applying is it is too slow at present to predict the deadly aftershocks that can happen a day or two after the first earthquake. But the good news is it will get better with time and with more data feed into the system the more information it will have to deal with the problem and slowly improve.