AI in connected cars could ease rush hour congestion, study shows

US researchers say they have successfully used artificial intelligence (AI) in a fleet of wirelessly connected cars to help ease peak-hour traffic on major roads.

In an experiment in Nashville, Tennessee, 100 test vehicles sent traffic messages back and forth on Interstate 24 (I-24).

They modified adaptive cruise control to use artificial intelligence to react to overall traffic flow to reduce so-called “phantom” traffic jams for no apparent reason.

The researchers are still examining the data, but say the experiment was a success. In addition to reducing driver frustration, less stop-and-go driving also means saving fuel and reducing pollution.

Daniel Walker, a professor of civil and environmental engineering at Vanderbilt University and part of a group of engineers and mathematicians from universities across the United States who have been studying phantom traffic jams, conducted a simple experiment in Japan more than a decade ago, shows how they develop.

Japanese researchers put about 20 human drivers on a circular track and asked them to drive at a constant speed.

It didn’t take long before the traffic went from a smooth ride to a series of stops and starts.

“The ghost traffic jams are caused by drivers like you and me,” Professor Walker said.

He explained that if a person hits the brakes for some reason, it takes a second for the person behind them to react and has to brake harder. Then, the next person has to brake harder. The wave of braking continued until many cars came to a halt.

However, as traffic cleared, drivers accelerated too quickly, causing more braking and another jam.

“We know that a car that brakes suddenly can have a huge impact,” Professor Walker said.

Experiments have shown that a few cars moving slowly and steadily can also have an effect — and more so.

The experiment used 100 vehicles that cycled a 15-mile stretch of I-24 each morning from 6:00 am to 9:45 am.

Assuming that if 5 percent of the cars on the road acted together, they could reduce the occurrence of phantom traffic jams, the researchers equipped those 100 cars with wireless communications, sending traffic information back and forth.

They also take advantage of adaptive cruise control, which has become an option on many new cars.

The technology lets the driver set the car to cruise at a certain speed, but the car automatically slows and speeds up as needed to maintain a safe distance from the car in front.

In the experiments, the adaptive cruise control system was modified to use artificial intelligence to react to the overall flow of traffic — including what’s happening ahead.

Professor Walker said there are two levels to the car’s decision-making. At the cloud level, information about traffic conditions is used to develop an overall speed plan. That plan is then broadcast to the car, which uses an artificial intelligence algorithm to determine the best course of action to take.

Researchers were able to assess the impact of connected cars on morning traffic flow using a special 4-mile stretch of I-24 equipped with 300 pole-mounted sensors.

The experiment is a project of the Circles consortium, which includes several automakers and the U.S. Department of Energy and Transportation. Other principal investigators are based at UC Berkeley; Temple University; and Rutgers University-Camden.

Liam Pedersen, vice president of research at Nissan, a partner in the Circles consortium, conducted the experiment in Nashville last week. One of the exciting things about it, he said, is that it builds on technology already in many new cars.

“It’s not on autopilot,” he said. “That’s something we’re going to realize pretty quickly.”