On Google Maps, artificial intelligence to predict traffic jams

A traffic jam on the weekend of August 1, near Chasse-sur-Rhône.

-

A new Google Maps update has just been announced.

It aims to improve the prediction of traffic jams and therefore estimate with more precision the journey times of users of the GPS application, reports

.

To do this, Google has decided to use a new artificial intelligence developed by Deepmind, a subsidiary of the parent company Alphabet.

“Our ETA forecast (estimated time of arrival) already has a very high accuracy bar.

We note that our forecasts have always been correct for more than 97% of trips, ”says Johann Lau, product manager at Google Maps.

The impact of the coronavirus on the calculation of journeys

According to Deepmind, this partnership will make it possible to improve the estimated arrival time by 50% in certain large cities such as Sydney, Sao Paolo, Berlin and even Jakarta.

Due to the current health crisis, Google Maps has had to change the way it calculates travel time.

The app now uses data that is two to four weeks old.

Indeed, it has become more difficult to predict traffic because of the sudden containment and deconfinement measures.

Beyond the traffic, Google indicates that it takes into account the quality of the roads or their size or their directivity to choose the route to be offered to users.

Economy

Taxes: Artificial intelligence helps make tax audits more efficient

High-Tech

Television: AI identifies which people and topics are most common

• High-Tech

• Application

• Traffic

• Artificial intelligence

• Traffic jams

• Google maps