When unexplained traffic jams happen, says an MIT Computer Science and Artificial Intelligence Laboratory (CSAIL) study, you can probably blame tailgaters. The researchers say that if drivers kept an even distance between cars rather than driving too close to the vehicle in front, traffic flow would remain even. This "bilateral control," could double the speed of the average vehicle on busy highways.
"We humans tend to view the world in terms of what's ahead of us, both literally and conceptually, so it might seem counter-intuitive to look backwards," says MIT professor Berthold Horn, co-author of the article. "But driving like this could have a dramatic effect in reducing travel time and fuel consumption without having to build more roads or make other changes to infrastructure."
According to MIT, the average urban commuter in the US spends 38 hours per year stuck in traffic, wasting 19 gallons (72 liters) of fuel. The costs of traffic congestion in the US are estimated to be $121 billion per year, or $820 per commuter, and use a total of 2.9 billion gallons (11 billion liters) of fuel. This does not include accidents caused by traffic disruptions.
The CSAIL study, published in the journal IEEE Transactions on Intelligent Transportation Systems, concedes that changing driver behavior would be a long, not very rewarding process. But what could change with relative ease is how adaptive cruise control systems found in most cars operate.
Adding rear radar sensors to a car with adaptive cruise could allow the vehicle to work to mediate its average distance between other vehicles, both ahead and behind. If just a few vehicles were outfitted and working this way, Horn says, the improvements to traffic flow would be immense.
The mathematics behind the idea are simple and can be likened to a vehicle's shock absorption system. When springs are placed between cars, as they are between the wheel and the vehicle's chassis, they flex and extend to continually maintain an equilateral distance between the vehicle and the wheel as a self-contained system. The CSAIL proposal is similar with a vehicle in question, using adaptive cruise control, maintaining equal distances between the car in front and the one behind through speed adjustments.
This ideal is very different from what is the norm in most thinking about traffic, especially by those stuck in it. Drivers (and, consequently, vehicle control systems) tend to be looking ever forward, responding only to what's ahead and largely ignoring what's behind. Thus, in stop-and-go or slow-and-go situations (traffic jams), each vehicle reacts to the vehicle in front, causing intermittent slowdowns or stops (jams) in wave-like patterns. When vehicles are working to maintain equal distances both from the car in front and the vehicle behind, the MIT paper contends, these wave patterns are minimized and traffic flows more smoothly.
And thanks to funding from Toyota and others, Horn is looking to undertake simulation testing in the future to determine if his method is indeed both faster and safer for drivers.
This latest study follows on from a more limited study of bilateral control conducted by Horn and published by IEEE in 2013. Another study by an electrical engineer, William J. Beaty, noted traffic patterns he called "Traffic Waves," outlined in a paper in 1998. Other traffic theories have been proposed, many based on fluid mathematics. The National Science Foundation has granted collaborative research, often based at MIT, on these studies.
More research has confirmed other traffic-busting ideas such as lane splitting for motorcycles and how artificial intelligence systems and self-driving vehicles could cope.
What's different about this latest study from MIT's CSAIL is its focus on solutions. Looking at both the micro- and macro-mechanics of traffic flow, Horn and Wang have proposed a relatively simple answer that can be applied using current technology.
Adding rear-facing radar to adaptive cruise control and programming the vehicle to maintain equal distances front and back, when possible, would go a long way towards resolving many tailgating-caused traffic jams. Even if, the researchers contend, only a relatively small percentage of vehicles on the road are using this technology adaptation.
Source: MIT CSAIL
