If you want to find the smartest drivers in the world, you need to head for the home of the US car industry. Just outside Detroit, lies the town of Ann Arbor, Michigan. The drivers there are not any more intelligent than other parts of the world, despite it being a famed college town. However, their cars are.
That’s because the roads of Ann Arbor are now home to a fleet of several thousand cars that constantly “talk” to one another. The scheme, known as the Safety Pilot Model Deployment project, offers a potential blueprint for the future of road transport. Like many projects it aims to cut congestion and make the road network more efficient. But this vision of the future is missing one thing: crashes.
“The idea behind it is to improve safety by an order of magnitude,” says Dr Peter Sweatman, director of the University of Michigan Transportation Research Institute.
Although fatalities and serious injuries from road crashes have levelled off in the US, they still account for more than 32,000 deaths each year, according to the US department of Transportation. Worldwide, the World Health Organization estimates more than one million people are killed in road accidents every year.
“This is an attempt to categorically avoid crashes, by having very robust communication between vehicles,” says Dr Sweatman.
The $22m project started in late 2011, but is on course to reach a major milestone in the coming days, when the project’s 3,000th connected car will pull out on to the streets of Ann Arbor. Each is fitted with wireless devices to allow them to talk to each other and to other parts of the road infrastructure, such as traffic lights. The technology is known as ‘Vehicle to Vehicle’ (V2V) and ‘Vehicle to Infrastructure’ (V2I) communication and allows the driver to send and receive information to and from the environment around it.
‘Real world experiment’
For example, the V2V allows the car to communicate with all other vehicles within a range of about 200m to 300m. They all share data such as their location, speed and direction, that is then assessed for safety risks, and if appropriate the driver is warned of any danger. For example, drivers will get an audible warning if they try to change lane with a car in their blind spot, or if the car in front of them brakes hard and the driver doesn’t seem to notice. The car can also give warnings at blind corners and junctions that another vehicle is about to pull out. Because the communications are done by radio signals, they do not need direct line of sight, unlike a driver’s eyes.
The V2I system can also communicate with other parts of the road and traffic infrastructure. For example, there are 29 intersections in Ann Arbor that have been instrumented, so that researchers can direct the lights in order to make the traffic move more efficiently.
“We’re collecting data from these intersections, and that is being back-hauled to the city of Ann Arbor through their fibre optic network so we can collect data, and make sure the signals are all being received and processed correctly,” Dr Sweatman explains.
To be clear, the technology fitted to the vehicles does not turn them into autonomous, self-driving cars, championed by the likes of Google, but they do represent an important step towards them.
“It’s hard to see how you could have automatic vehicles without them being connected,” says Dr Sweatman. “It’s technically possible, but in terms of a realistic deployment autonomous vehicles will need to be connected.”
Other vehicle interaction systems are in development, and what is learnt in this huge, real world, experiment will help inform the direction they take.
“This is fundamentally aimed at safety, but clearly with a platform like this you can create many other applications,” says Dr Sweatman. “There are many other potential benefits in mobility, environment, emissions, energy efficiency.”
For example, the team envisages a system that monitors the brakes of all of the cars. If the traffic management system detects that all of the cars are braking at the same spot, it could indicate a problem with the road that needs to be fixed by a maintenance crew
Alternatively, if for some reason there was an accident, the system could be used to immediately alert the emergency services, divert other drivers to alternative routes and ensure that the roads are clear for emergency vehicles.
Currently, the majority of the cars fitted with the technology are owned by residents of Ann Arbor, who have volunteered to join the programme. People take their cars in, just like a service visit at a dealership. It takes less than an hour to install the two antennas for GPS, one antenna for communications, the driver interface, and a black-box computer to link it all together.
So far, about two thirds of the drivers selected to take part in the trial are parents from Ann Arbor schools. They were deliberately chosen to ensure there are as many vehicle “interactions” as possible, because they tend to drive their kids around a lot, and tend to congregate in the same places. More than 60 vehicles, including buses and trucks, have also been donated by various car manufacturers.
“We believe that at any given time there will be 5 to 10% of these vehicles in the traffic stream,” says Dr Sweatman.
However, this number could rise dramatically if the trial – which is currently due to run until 2013 – is successful. Then, the technology could be rolled out globally. With the ability to save thousands – perhaps millions - of lives, and reduce congestion and pollution at the same time, Ann Arbor could be the blueprint for the future.