We've all been there. Stuck in a city-centre traffic jam, a sea of red brake lights ahead. Grumpy and stressed, you edge forward, silently (or occasionally, loudly) cursing the other drivers on the road.
When asked, somewhere between 80 and 90 per cent of drivers believe that their skills behind the wheel are above average. Clearly, lots of drivers are wrong. But the idea that traffic jams must be the fault of someone else is a pervasive one, and it's reflected in the language we use to discuss them. We say things such as, "Oh, the traffic was terrible," or, "The roads were so busy" - as if the jam is a separate entity to the drivers caught up in it.
The science behind how and why traffic jams form tells a very different story. And it is one that's being collectively written by researchers from a diverse range of fields. From molecular ecology and human behaviour to network science and urban planning, there are thousands of people trying to understand traffic and find new ways to keep it moving.
Road congestion can be caused by any number of factors - bad weather, an accident, roadworks or just by too many vehicles competing for too little space. But there's also the "jamiton" or "phantom traffic jam", where, for no discernible reason, traffic builds up and then eases.
A now-famous video of how these types of jams form shows 22 cars being driven on a closed track. The driver of each car was instructed to get up to 30kph and maintain that speed at a safe distance from the car in front. But the system broke down very quickly, with some cars left at a standstill while others were accelerating. The reason is simple - people have trouble maintaining a constant speed. Say one driver finds themselves driving just slightly above the speed limit - to correct for it, they then tap on the brakes. The car behind then overcompensates for this sudden braking, as does the car behind that one. This causes a start-stop shockwave that travels backwards through traffic.
We usually associate the concept of collective behaviour with the natural world more than the urban jungle. A paper published by German scientists in 2015 looked specifically at traffic flow in ant colonies. Black-backed meadow ants construct and maintain permanent roadways not unlike our own - a fixed width and a smooth surface, clear of obstacles. By observing ants using the route, Christiane Hönicke and her colleagues could investigate the ant etiquette involved in the rapid flow of traffic in and out of the colony. Surprisingly, they found that, as the trail got more crowded, the ants sped up. In fact, they increased their speed by about 25 per cent as the density doubled.
At these higher speeds, collisions were more frequent between ants - possibly not a situation that human drivers should be emulating. But another factor in ant traffic flow was the evolution of "lanes" as the route got busier - something that has been observed by mathematical modellers in India too. Most researchers also agree that individual ants tend give each other a lot of headway. This gives them more time to react to any incidents, reducing the risk of kick-starting a jamiton. Ants could teach us a thing or two, but changing driver behaviour on the roads can be a challenge. For example, dynamic late merge, also known as the "zipper system" has been shown to greatly reduce congestion when merging two lanes into one. But despite this, many drivers still opt for the more polite early-merge option.
Marta C González of MIT's Civil and Environmental Engineering Group believes that we could all benefit from taking a less selfish approach to driving. Using smartphone data, she showed that taking just one per cent of cars off the roads from specific neighbourhoods in Boston and San Francisco could reduce travel time for all other drivers in those cities by up to 18 per cent. In early 2016, González found that in cities including Rio de Janeiro and Lisbon, even if a small number of drivers took a slightly longer route, the total time lost to congestion could drop by 30 per cent.
Infrastructure has a role to play too, including traffic lights within cities. They use sensors embedded in the road to continuously feed information on traffic flow back to a central control centre. But cities such as London are making lights smarter. There, thermal cameras monitor the number of pedestrians and cyclists at certain road junctions and adjust the "green-light time" to give them an official head start. In August, Audi announced that their new Q7 and A4 cars will be able to communicate with smart traffic lights, providing a green light countdown for drivers.
Longer term, the biggest challenge facing traffic managers will be the mix of transportation on the road - namely, a growing number of autonomous vehicles, surrounded by many human-controlled ones. A full move to driverless cars would have a major impact on road infrastructure too. Because they can continuously communicate with each other, driverless cars could potentially speed safely through junctions, removing the need for physical traffic lights. Where will that leave pedestrians? Well, that's a question that designers don't yet have an answer to.
This article was originally published by WIRED UK
