Google’s maps and navigation have become an unavoidable travel companion for many, whether it’s driving through unfamiliar places and cities or driving daily routes as Google notifies them of traffic jams, car accidents, and more.
While using maps is very simple, calculating movement routes, estimating travel times, and predicting traffic so drivers can choose the fastest route is an extremely complex task.
For all those who want to know how Google manages to calculate the best routes with very high accuracy (for more than 97 percent of trips), warn where traffic jams will be, and predict travel time, the company published an interesting article on its official blog in which they discovered some hitherto unknown details. In all of these calculations, Google uses machine learning and traffic data it has collected over the years in its maps.
These data are analyzed and show what the average condition of the roads is at a particular time of day. How traffic is calculated is explained in the example of moving on Highway 298 in California. On it, between 6 and 7 in the morning, the vehicles usually move at a speed of about 65 miles per hour, but in the afternoon, the speed on that highway, due to heavy traffic, is only 15-20 miles per hour. Google combines this data, based on years of road traffic analysis, with a real-time situation using a machine learning method to predict the duration of the trip.