How connected cars will optimize traffic flow

Algorithms along with ubiquitous automobile data connections will make traffic jams a thing of the past if scientists get the calculations right.

It has been hard to miss the buzz circulating that the so-called "driverless car" is just around the corner. Futurists have even speculated that one day humans won't be allowed to drive cars—it's just too dangerous, and computers will do a better job.

Well, the driverless car is clearly on its way, and I, for one, am looking forward to creeping along in one. But in the meantime, we're just going to have to make do with "connected cars" for our entertainment.

And connected cars have a lot to bring to the future road party. Tangled traffic at intersections will be on its way out if scientists at the Urban Dynamics Institute (UDI) at the Department of Energy's Oak Ridge National Laboratory get their algorithms correct.

Computational framework

The researchers are trying to figure out how to reduce travel time and fuel consumption by getting vehicles to talk to each other and communicate with traffic controls, such as traffic lights.

Location, speed, and destination should be exchanged with traffic controls and other road users in order to create instructions for drivers. The researchers said that this can eliminate stop-and-go driving.


Now, you might think you've heard this before. And, in fact, you have. Many parts of the world have experimented with traffic management through road sensors, cameras, and so on over the years. I can remember hearing about this stuff when I was a teenager.

The difference, though, is that in those days we didn't have the sophisticated algorithms that we have today, and we also didn't have the connected cars. Drivers have had cellphones for a long time, but they haven't been hooked up to the workings and sensors of the automobile.

Cars are about to become even more connected. European regulators will likely force automakers to build cars with emergency call technology by 2018.

European parliamentarians want cars to dial a Euro-wide 112 emergency number automatically in the event of a collision. It will work a bit like GM's OnStar system, which provides location through GPS to an operator.

So all cars are ultimately going to have SIM cards and cellular radios in them anyway.


Future connected cars will tell drivers the "optimal speed, the best lane to drive in, or the best route to take," Andreas Malikopoulos, UDI deputy director, said in an Oak Ridge National Laboratory website article.

The first step is building the decentralized control algorithms, which determine how vehicles will interact.

No central control center

Oak Ridge's system intends to abandon the idea of the central control center—something usually seen in city-wide traffic management. It says that these centers generate too much data, and that it's not realistic to expect all vehicles in a city to communicate that much information all at once.

The system suggests that vehicles communicate among each other locally and across a city.

In addition to communications, the algorithms will also include analysis of traffic patterns and conditions. Large-scale data compiled through simulations in real urban areas will be used to make predictions, like when school zones are busiest—in the morning and middle of the afternoon, for example. That kind of information will be pumped into the system.


And unlike the "driverless car" —of which this kind of algorithmic traffic management is a precursor—in this case individualized instructions will be displayed for the driver to act upon.

Which begs the question—what happens if the drivers don't drive how the algorithms tell them to? Chaos, right? Well, the UDI scientists have thought of that, because one element that they're looking at including is digital ticketing.


Copyright © 2015 IDG Communications, Inc.

The 10 most powerful companies in enterprise networking 2022