Smartphones connect over peer-to-peer networks with ease. In fact, with the latest Android phones, you need only touch the backs of two handsets together to instantly start sharing data. With your car, the only peer-to-peer experience occurs when you touch the bumper of a Ford at your local supermarket.
Fortunately, that's about to change, because an important car-to-car roadway test has now started in the US.
The benefits of peer-to-peer car comms are obvious - you'd get realtime traffic, accident and weather updates, etc, live from cars all around you, enabling your sat nav to avoid tailbacks, dodge accidents and generally get you to your destination as fast as possible.
In a world where roads are already overcrowded, this kind of tech will be vital in the years to come, in order to keep traffic flowing at optimum efficiency. In theory, this will eventually lead to car sat nav systems that plot and show you the locations of every car on the road in your vicinity.
Of course, the Car 2 Car Communication Consortium in Europe has experimented with wireless transmissions between cars for some time. In late August, a new project started in the US that seeks to find out what happens when thousands of cars communicate with each other over a full year.
The project, called the Safety Pilot Model Deployment, started on August 21 with 300 cars on a highway near Ann Arbor, Michigan. Eventually, 2,850 cars and trucks will participate.
The project is a collaboration between the US Department of Transportation (DOT) and the University of Michigan. For the first time, a major trial will show if car-to-car communication can work under real road conditions.
"This will demonstrate the technical maturity of the technology," says Don Grimm, a senior researcher at General Motors (GM). "While previous deployments were more experimental, this is the first time a trial is using production-like cars where regular drivers participate over a number of months."
Driving information exchange
GM is one of eight car manufacturers participating, having 'instrumented' eight cars in a fleet of 64 from other automobile makers such as Ford and Nissan.
The driver won't see any of the electronics but will use a dash interface which varies between car manufacturers in the trial. Other vehicles will use a device that is not embedded into the car but can still transmit and receive data.
The idea is to provide more information about other cars and vehicles in real time. For example, the driver might see an alert that another motorist had to brake suddenly several kilometers/miles ahead, that there is a traffic jam along one stretch of roadway, or that one vehicle has collided with another. Drivers benefit from the aggregated data and can plan to slow down or change their route.
Current vehicles use high-powered sensors to scan the surroundings and can sense an imminent collision, applying the brakes accordingly. For example, many BMW and Mercedes-Benz cars can brake to avoid a crash. However, these short-range systems only work 30m/100ft ahead of the car.
The new trial uses a wireless radio spectrum called Dedicated Short Range Communications (DSRC) in the 5.9GHz band over a range of about 180-210m/600-700ft. About 400 cars and trucks will be able to communicate with each other.
In a mass roll-out, a server infrastructure installed by city and US federal entities could handle communication between thousands of cars at a time. DSRC was first introduced in 1999, but the Ann Arbor test is the first to use the spectrum for thousands of vehicles.
The Ford Taurus Sho in the trial uses lights, tones, haptic sensations in the seat and lights in the side mirrors to warn the driver when there is a DSRC message. One interesting example of how the warning works differently from existing onboard sensors is when passing on a two-lane road.
Today, a sensor tells you that the car ahead is braking. But DSRC will enable us to know there is a car coming over a hill that we can't see yet, telling us to stay in our lane and wait until the car passes.
Mike Shulman, technical leader of Ford Active Safety Research and Innovation, told us that most of the drivers in the trial work and live in the same area.
The 400 car and truck sample size is a technical limitation that could be solved easily with more servers and more robust infrastructure, he says. As it stands, the 5.9GHz signal sends out seven isolated signals at 10MHz each (plus two guard signals) for a total of 75MHz, all allocated by the FCC in the US.
Are self-driving cars really the future?
Each vehicle sends out a signal 10 times per second, although Ford uses an algorithm that can detect when a car is parked and reduce the transmissions accordingly.
"Vehicle manufacturers and DSRC suppliers are working on communication channel congestion mitigation strategies, which will be able to handle large quantities of vehicles within communication range. Mitigation schemes are based on message priority, which takes into account the proximity and speed of other vehicles," says Nancy Wilochka, a spokesperson for the DOT.
To handle security, the cars transmit to a roadside station that confirms their location and broadcast credentials. In a wider deployment, these roadside stations would be scattered all over a city. The car-to-car communication would extend further to transmit data to more roadside stations, helping with traffic management and even traffic forecasting, road maintenance, and local regulations.
In-car tech: ultimate guide to infotainment
The Ann Arbor test points to a day when all vehicles communicate with each other and with the roadway. Like the peer-to-peer music networks we use today, drivers would know where other connected cars are on the road at all times. If one car makes a sudden swerve on the road to avoid an obstacle, a car following several kilometers away could see an alert about this obstruction.
"We have worked for a long time on passive safety, and we started putting sensors into cars to sense what is around the car to warn the driver to mitigate or avoid a crash. But vehicle-to-vehicle is a safety sensor to give the car a lot more information: path prediction, brake status, heading and yaw rate, and more. When you put that together you get a map of what is going on around you," says Shulman.
The next step after that, he says, is to consider more automated driving. A computer in a car can think faster and more logically than any human. That said, the technology is still in an early stage, as computers learn to think with the clarity of a human, telling the difference between a simple tree branch in the street and an animal that has scurried into the road.
"Warning functions are one thing, but the next step is to tie the DSRC messages to the car systems. If the driver does not react to a DSRC warning, the car could eventually steer out of a crash or apply brakes automatically. You start to see how there could be much more automated driving," he says.
Of course, with all of this technical investment and a year-long trial, you might wonder why car manufacturers are not thinking more about using mobile devices for peer-to-peer communication.
According to the DOT, the main issue is latency. DSRC is a more predictable technology, in that it only serves one purpose (safety warning) and can be scaled for city infrastructure. Smartphones can easily overload networks and cause congestion, something that vehicle companies want to avoid in more ways than one.