The Stack Archive

Smart driverless car system proposes to replace traffic lights

Fri 18 Mar 2016

Traffic jam intersection

A new traffic light-free system has been proposed which would see the century-old technology replaced by a capacity-boosting and environmentally friendly alternative.

In a cross-institutional study titled Revisiting Street Intersections Using Slot-Based Systems, researchers describe a system whereby driverless vehicles are assigned ‘slots’ when approaching an intersection. Each car knows exactly when it should arrive at the intersection and when it should travel across, adjusting its speed accordingly. As the slot system is based on mathematical modelling and communicates directly with the driverless vehicles, it obviates the need for traffic lights entirely.

The proposition would improve road capacity, since intersections would be able to manage double the amount of traffic. The researchers suggested that this capability would also help reduce emissions, as cars are able to complete their journeys faster.

Screen Shot 2016-03-18 at 11.39.00An effect known as the ‘slower-is-faster’ principle occurs when the slot-based technique slows down the vehicles but increases the throughput, leading to an overall reduction on the average delay. Evidence of the effect, the paper suggests, has already been observed in traffic, pedestrian movement, production, and logistics.

The team argued that intersection management technology demands urgent attention from engineers and planners as they prepare to transition towards smarter city infrastructure that will be in operation for several decades to come. They explain that the intersection is a critical area to target, as a breakthrough could result in a major positive effect filtering through the entire network level. ‘The doubling of bottleneck capacity, as promised by [slot-based intersections], has the potential of significantly reducing overall congestion and improving the stability and predictability of traffic,’ the paper outlines.

Further work will look at scaling the project to a network of road intersections. Such scaling would be more efficient than conventional traffic, since all of the system algorithms require ‘minimal computational effort.’

Google is contributing to another collective, this time with the U.S. Department of Transportation, regarding the development of future traffic management technologies at its Sidewalk Labs. The tech giant is currently developing a smart data platform called Flow to improve traffic congestion, urban pedestrian transit, and driverless systems.

In an announcement yesterday Sidewalk CEO Dan Doctoroff said: “I think we could all agree just from our daily lives that there may be no bigger problem than the problem of congestion…We know we’re not going to build more roads; it’s hard to invest meaningfully more in mass transit…We have to do better with what we have.”


news research self-driving cars smart transport
Send us a correction about this article Send us a news tip