My friend was complaining to me the other day and asking why as soon as it gets to peak hour, people forget how to drive?
There are two answers; one involves just agreeing because you want to support your friend; the other is rather technical and involves a little bit of traffic engineering. But both answers come back to one thing, the humans are not as good at driving as robots are when things get busy and vehicles are closely following each other (or as good as robots could be in the future anyway).
On the bahavioural side of things, humans can get distracted singing along to their favourite song, checking their phone and talking to their passengers but robots don't have the same distractions. Also, all humans drive slightly differently and have different priorities. Some are in a hurry, some are not. When things get busy, it's not that people forget how to drive, it is just that all drivers are in it for themselves (the tragedy of the commons) and there becomes a point that humans can't drive the way they want to, which can be quite frustrating.
A vehicle driving along a stretch of road needs a certain amount of space for safety and depending on the speed, it could be quite large, like when you are travelling on a freeway at 100km/h, or it could be quite small, like when you are following someone closely at walking pace in peak hour traffic. Ultimately though, once all the space on a road is taken up by vehicles, things would totally break down as shown in the picture above which I took at the Arc de Triomphe in Paris during the peak hour. It took a couple of Police around 10 minutes to sort the situation out by slowly moving vehicles out of the way. If robots and computers were driving cars, traffic has the potential to operate better for much longer during peak hours, making the entire network and all trips on average shorter than they would be if humans were driving like they currently do.
It is pretty hard to explain in words and not much fun or interesting using maths, but the video below does a pretty good job at showing how things can break down when too many vehicles are competing for space and humans are driving the vehicles.
If you are a Mythbusters fan, there is a full episode where they investigated a number of traffic myths which included replicating the experiment in the video below. You can watch the full Mythbusters episode here.
Looking at the video, you can see that while the gaps between cars starts off evenly, it quickly develops into a peak hour traffic jam, just like what drivers are caught in every day on busy stretches of roads and at busy intersections all around the world. Human drivers get distracted and when changing lanes, if drivers hit their brakes too hard, that can cause a traffic jam in itself.
Back in June 2014, we looked at the research into driverless cars being done by Google but there are many players currently undertaking research into driverless technology and connected vehicles:
- Nissan and NASA have teamed up for the next five years and aim to have an autonomous car that can drive in nearly all situations by 2020 [1]
- Ford has recently opened a new research facility in Silicon Valley which will be run by a former Apple Engineer [2]
- Mercedes Benz has recently released a unveiled a new driverless concept car [3]
- An Audi A7 drove itself from Silicon Valley to to the Consumer Electronics Show in Las Vegas January 2015 [4]
- BMW has announced it has developed a valet style self-parking feature [5]
- Honda showed off its driverless technology at the 2014 ITS World Congress in September 2014 [6]
It appears though that not all car companies are actively developing fully autonomous vehicles with Toyota saying it will instead focus on technology and safety features and their connectivity with the driver [7].
So back to the problem at hand, peak hour congestion and how can connected vehicles and driverless technology help? On daily commutes it would be handy to let the car drive for you, but that is not going to be happening on a large scale any time soon despite car manufacturer aims of having fully autonomous vehicles on the road by 2020. What is more likely are a series of improvements in vehicle to vehicle communication which will allow for better traffic flow and over time, once the technology gains acceptance, autonomous vehicles will very likely become a common occurrence on our roads.
Already available in a number of vehicles is adaptive cruise control. This type of technology would allow the cars on the roundabout shown in the video above to stay at the target speed of 30km/h and if all vehicles on a busy freeway were using that type of technology, traffic would be able to flow better.
Other potential improvements include:
- Advanced braking warning for cars ahead that are slowing down but cannot be seen by the driver.
- Connection with traffic signals to provide for better signal coordination and capacity of intersections.
- Advanced safety warnings, for example, not to overtake a slow moving vehicle.
- Warning of weather related issues such as black ice or water over the road.
- Notification of nearby emergency vehicles.
The United States Department of Transportation has produced a video which looks at some the future possibilities for connected vehicles and in reality, some of the possibilities are not very far away.
With the pace of computer development and technological advancement in the last 10 years (the iphone was released in 2007), it is hard to predict will happen in the next 10 years. But if the pace of development continues as it has in the last few years, it is possible to imagine a future where humans are not driving the cars that they own, at least in some situations. But until then, people will have to just be patient with peak hour traffic and remember that not all drivers are created equal.
[3] https://www.mercedes-benz.com/en/mercedes-benz/innovation/research-vehicle-f-015-luxury-in-motion/
[4] https://www.youtube.com/watch?v=NiiGmji0bFk
[5] http://www.themotorreport.com.au/60462/bmw-to-debut-self-parking-tech-at-ces-2015
[6] http://www.honda.com/newsandviews/article.aspx?id=7926-en
[7] http://www.pcmag.com/article2/0,2817,2468022,00.asp
Author: Matthew Houlden
All references last accessed 27/1/15
