The Silicon Review ^Asia

A team of scientist from Australia has developed software by using AI and mathematics. The software can detect and prevent landslides weeks before it occurs.

The scientists utilized radar data from various mining companies to create the software. After collecting the data, they searched for patterns to find out the movements of unstable locations. The algorithm was developed with the help of data from a landslide-prone Italian volcano.

The software also has the potential to include data about other landslide risk factors, like erosion and rainfall. It will make the targets even more accurate. Radars on the ground satellites or even in drones will provide the data for monitoring.

According to the team, currently, a lot of prediction on landslides is based on People’s instinct on the location. The team doesn’t rely on the instincts. The data set is very huge and building the software is a big achievement for the team and its mainly finding a needle in a haystack filled with needles. The main goal is not only to find too many patterns but detecting the correct pattern out which can give a clue on the location where the landslide may occur in the future.

It is believed that the scientists are also seeking ways to test their tools for structural health monitoring, predicting the collapse of structures like bridges or dams.