If you're comfortable with the present green, amber, and red traffic lights, be prepared to get uncomfortable. New research suggests that adding a white light will speed up traffic and improve safety for both cars and pedestrians.
Red for stop, green for go, and amber for prepare to stop (or stomp on the accelerator, if you're like some drivers). It's been that way since the 1920s and is even codified by an international treaty to make sure accidents don't occur because a London driver doesn't understand signals in Tokyo.
Over the past century, the technology has been undergoing constant refinement. Electric lights replaced semaphores. Computers allowed multiple lights to be linked, and sensors allowed them to analyze traffic flow and adjust timing for maximum flow.
Systems have been developed that control lights for whole cities and digital models of traffic flow have been honed to a fine art.
The emergence of autonomous vehicles (AV) has introduced a new twist. Instead of cars being regarded as little more than wooden blocks being pushed around on a map, AVs not only have the ability to drive themselves, they can also communicate with one another as well as a central traffic computer system.
What this means is that the AVs become part of the traffic control system itself and can work together to improve traffic flow by a considerable margin. A team of researchers from North Carolina State University have suggested that once a critical number of AVs are on the road, a fourth, white traffic signal light can be added to the traditional trio. Their computer model worked very well at speeding up moving through intersections, but there was room for improvement.
"Our earlier work introduced the idea of a fourth traffic signal called a ‘white phase,’ which taps into the computing power of autonomous vehicles in order to expedite traffic at intersections – but we had not yet incorporated what this concept would mean for pedestrians," says Ali Hajbabaie, an associate professor of civil, construction and environmental engineering at North Carolina State University. "We’ve now expanded our computational modeling to account for foot traffic, and the results are extremely promising for both pedestrians and vehicles."
As in the previous study, the AVs act as shepherds for the human-controlled vehicles in their lanes. When enough AVs are present, the signal turns to white, telling the human drivers to follow the AV or the other car ahead of them. The more AVs, the faster the flow. The difference this time is that pedestrians are now incorporated into the model.
What the team found was that even with foot traffic present, the general flow still improved for both vehicles and pedestrians by over 25 percent.
Of course, there's still a long way to go before such a system becomes practical. Aside from getting enough AVs on the road, there's also the problem of installing enough of the four-light signals. Even then, there's getting human drivers to trust the new white signal and to modify signals so that they will be meaningful to pedestrians, such as introducing blinking green lights or something more immediately legible.
"We are currently setting up a physical testbed that will allow us to experiment with this concept in the physical world – not just in a computer model," said Hajbabaie. "However, the vehicles we are using in the testbed are small enough to hold in your hands. This will help us identify challenges in implementation without the expense – and safety risk – involved with using full-scale vehicles. In the meantime, we are open to working with industry and research partners to explore ways to move forward with these technologies."
The larger question, of course, is how to fit the new light into the songs we sing to our kids. "Stop, says the red light. Go, says the green. Wait, says the yellow light, sitting in between." Not an ounce of fat on that. "Follow the autonomous car in front of you, if there is one, and cross the road too if you want to, or otherwise just stay where you are, says the white." Hmm, needs work.
The study was published in Computer-Aided Civil and Infrastructure Engineering.
Source: North Carolina State University
