In the forefront of cities experiencing congestion

Artificial intelligence is trying to solve the traffic problem in Los Angeles

11,160 sensors are deployed on the main road in California, USA.


Los Angeles, in the US state of California, is famous for its traffic jams, and it is at the forefront of cities that suffer from traffic problems and congestion in the United States, to the point that it is estimated that Los Angeles citizens spend 120 hours a year inside their cars in the midst of traffic chaos.

As part of efforts to address this problem, the Los Angeles transportation authorities - in cooperation with the Argonne National Laboratory of the US Department of Energy, and researchers from Lawrence Berkeley National Laboratory in the United States - developed an artificial intelligence system to help reduce traffic congestion in City streets, and the technology website "", which specializes in technology, reported that experts in supercomputers in the Argonne laboratory have created an artificial intelligence system that can accommodate traffic patterns in the city over an entire year, with data they obtained from 11,160 sensors. They are spread on the main road in the entire US state of California.

The artificial intelligence system was fed with this data in order to predict traffic movement and places of congestion, and average vehicle speeds on the roads, and the new system can analyze traffic data in the past hour, and then make predictions, within fractions of a second, about the places of traffic congestion during the next hour accurately. Extreme.

"Artificial intelligence techniques and supercomputers have been adapted to treat this serious problem," says a researcher in the field of computers in Arjun's laboratory, Prasana Balabrakash, adding that "this project requires huge amounts of data and super computing capabilities to deal with it", explaining that an ordinary desktop computer can That it takes three weeks to train the traffic program to predict traffic, while this process does not take more than three hours by a supercomputer.

He added that the previous traffic programs could handle data collected from 200 to 300 sensors, while the new system could deal with data obtained from more than 11 thousand sites, in addition to carrying out this task in record time compared to previous programs. .

Los Angeles citizens spend 120 hours a year in their cars in traffic chaos.

Follow our latest local and sports news, and the latest political and economic developments via Google news