Researchers from the Complex System Group at the Institute of High Performance Computing (IHPC), Agency for Science, Technology and Research (A∗STAR) have modelled a system which would enable cars to cross traffic intersections safely without the need for traffic lights.
In smart cities of the future, communication beacons would complement and eventually replace traffic lights. The beacon gathers and transmits data about the distance and approach speed of vehicles nearing the intersection. The information is received by the crs and fed into an algorithm which plots a safe course through without the car having to stop.
According to the researchers’ modeling, each car will be able to cross the intersection in its own virtual bubble of safe space, modulating its speed using adaptive cruise control. It utilises the concept of “adaptive repulsive force”. The closer two cars’ trajectory would bring them at an intersection, the stronger their repulsion and the greater the speed adjustment they make to pass each other safely.
The algorithm of the scheme is completely decentralised. The researchers found that the rules governing the necessary repulsion between vehicles are rather simple, resulting is a system that does not require much computing power at the beacon or in the vehicle itself.
For vehicles approaching the intersection in different directions, the algorithm imposes simple interactions between vehicles around the intersection, by defining specific conditions on the real-time basis, for which the involved vehicles are required to briefly adjust their dynamics. This leads to a safe, robust, and efficient self-organised traffic flow. The researchers also took into account the driver comfort level and studied its effect on control efficiency.
The algorithm worked effectively even for relatively complex intersections in the simulations. “In most cases, pre-emptive deceleration only slightly lowered the vehicle velocity, resulting in safe passage of each vehicle across the intersection without coming to a full stop at any point,” said Bo Yang, from A*STAR IHPC, who led the work.
No need for completely self-driving cars
An important consideration for the practical applicability of the traffic intersection control is the minimal level of intelligence the vehicles need to have for the system to work.
The proposed system does not require cars to be completely autonomous or self-driving. It only requires a majority of vehicles on the road to be equipped with a simple driver assistance system, which is able to brake and accelerate autonomously. Cars fitted with smart cruise control can already do that. The driver gives up control of the car’s speed through the intersection but remains in charge of steering.
The system’s other advantage is that it could be phased in gradually. It has minimal impact on conventional driving behaviour, and can coexist with the existing traffic light control. For initial implementation, traffic lights would still be needed to help older cars pass through the intersection. As smarter cars become more and more prevalent, the lights can switch off for more and more of the time until they are no longer required. And though the algorithm is using only basic vehicle intelligence, it is fully compatible with the futuristic scenario in which driverless vehicles dominate the road.