Telematics, V2V and autonomous vehicles

The US National Highway Traffic Safety Administration said it will decide whether to engage in rule making for vehicle-to-vehicle (V2V) communications in 2013, following its year-long model deployment, which kicks off in Ann Arbor, Michigan in August.

V2V, which would allow cars to transmit brief messages to each other via embedded short-range modems, could reduce collisions. However, to get this benefit, a significant number of vehicles on the road would need to have the technology installed. How many? That's still to be determined. (For more on V2V and vehicle-to-infrastructure (V2X) technology, see Special report: Telematics and V2V/V2X technologies.)

Meanwhile, Google's autonomous Prius has been cruising the streets and highways of the San Francisco Bay Area for two years. The State of Nevada Department of Motor Vehicles is ready to license autonomous cars, and Continental aims to make its experimental Passat one of the first to be licensed. Toyota and Ford also have experimental vehicles that can drive themselves, thanks to arrays of sensors, scanners, and radars.

Industry boosters claim we'll see at least semi-autonomous, non-experimental cars on the roads in three to five years. Will the high speed of autonomous vehicle development make V2V irrelevant?

On the road today

Most auto manufacturers have cars on the road today with advanced driver assistance safety (ADAS) features, such as forward collision warnings and blind spot warnings. It's a relatively short hop to greater autonomy. For example, Continental's experimental Passat is equipped with off-the-shelf or close-to-production technologies for monitoring a vehicle’s immediate surroundings and actuators. (For more on ADAS, see V2X telematics: Taking ADAS to the next level and Telematics: Backing up on backup cameras.)

Says Christian Schumacher, director of advanced engineering for Continental, "We are undertaking feature-by-feature development, rather than releasing an autonomous car." The company's philosophy is to build on its already-released systems for low-speed features like parking assistance, moving up to speeds of 25 or 30 miles per hour. "The next step we'd like to work on and introduce to the market is traffic jam assist," Schumacher says. Traffic jam assistance would provide for highly automated driving; the car would handle acceleration, braking and steering in stop-and-go traffic.

If a vehicle can do this on its own, does V2V become irrelevant?

John Kenney, senior research engineer at Toyota InfoTechnology Center, USA, points out that OEM-initiated ADAS let individual companies innovate at their own pace: "The big difference is that ADAS is something an automaker can do on its own. What one decides to deploy doesn't depend on what anyone else does, and they don't have to have same thing in each model. They can be pretty agile."

ADAS as short-term solution

On the other hand, for the automaker, dedicated short-range communications (DSRC) could eventually be a cheaper solution than ADAS. Instead of installing the cameras, radars and sensors needed to ‘see’ around the car, as well as the software and processors to make sense of that data, V2V would let automakers install the cheaper DSRC modems that provide a steady stream of standardized data, with maybe a couple of short-range radars to supplement.

The problem with relying on DSRC, of course, is penetration. Says Steven H. Bayless, seniordirector, telecommunications and telematics at the Intelligent Transportation Society of America, "If it is true that V2V is a cheaper alternative to ADAS systems that depend on sensors like radar, that's great. But you still have to wait until there's significant penetration on the roads."

Certainly some applications, such as forward collision warning, don't require high penetration at all, he notes. If the car ahead of you has a DSRC beacon, you get the benefit; if not, you rely on your reflexes, like always.

On the other hand, for applications such as blind spot detection, he thinks penetration would need to be close to 100 percent in order to get the safety benefits. "It's on an application-by-application basis," Bayless says. "Applications have to be treated differently from the underlying DSRC technology."

Better together

The answer to whether we'll rely on V2V or autonomous driving may depend on who you're talking to—ADAS provider or autonomous driving researcher. But most agree with Vann Wilber, senior partner of Global Technical Policy Associates and policy program manager for the Vehicle Infrastructure Integration Consortium, who says, "I think the two will coexist; there will be a convergence of some kind. I see them being complementary and maybe one becoming more sophisticated, and the other not having to be so sophisticated."

He points out that, for example, a car's forward collision-warning system can't see two or three cars ahead of the car in front; V2V could expand that visibility. Or, a car might have a forward-facing camera but not one to see what's coming from the side.

A full 360-degree sensor system, such as that employed by Google's cars, Wilber says, is "totally self-contained and very high-precision; that is, very expensive technology with a lot of redundancy, like a space shuttle almost." Layering on the communications link could make it even better.

Continental's Schumacher agrees: "I see vehicle-to-vehicle and vehicle-to-infrastructure as a key enabler for automated driving. The communication gives us more data points for automated driving, and the communications part might be one of the key enablers to approach the next level of automated technology."

Kenney argues that, while it seems that development of V2V communications lags ADAS innovation, plenty of progress is being made. "DSRC will only happen if we can do it in an interoperable way,” he says. “By necessity, we're joined together in moving this forward. It might look like it's moving slower but we … have overcome a lot of the challenges that were there at the beginning."

Susan Kuchinskas is a regular contributor to TU.

For more on V2V and vehicle-to-infrastructure (V2X) technology, see Special report: Telematics and V2V/V2X technologies.)

For all the latest telematics trends, join the industry’s other key players at Insurance Telematics USA 2012 in September in Chicago and Telematics Munich 2012 on October 29-30.

For exclusive telematics business analysis and insight, check out TU’s reports on In-Vehicle Smartphone Integration Report, Human Machine Interface Technologies and Smart Vehicle Technology: The Future of Insurance Telematics.

Autonomous Vehicles