The autonomous car: The road to driverless driving

No longer the stuff of science fiction: a car that's a safer driver than you are.

While major automakers unveil their visions for the autonomous car, and authorities move toward licensing them, the latest advanced driver assistance systems (ADAS) make it clear we're halfway there already.

So-called driverless cars, tricked out with cameras and sensors, are great news for automotive suppliers of all manner of sensing technology. But they could be profoundly disruptive for automotive OEMs and other stakeholders, reshaping everything from how cars are owned to how we move through urban settings.

At the moment, there are two parallel roads to driverless driving: One heads straight to a self-driving vehicle, with suppliers working on complete autonomous driving systems they hope to sell to auto makers. The other road involves automakers and parts suppliers working on incremental advances in ADAS that could eventually fuse into an autonomous system.

Look, no hands!

The leader of the autonomous pack is Google, which has garnered kudos for its highly visible self-driving Priuses. Bosch and Continental Automotive Systems, as well as Audi, also have fully autonomous cars on the road, while Continental recently announced it would work with BMW on autonomous driving systems.

The trick for these companies is to get the size and price of the various sensors down low enough to make these systems feasible for consumer vehicles. Google's and Bosch's cars, for example, depend on roof-mounted, spinning LIDAR systems that can cost from $25,000 to $70,000 each.

Continental's vehicle sticks to sensing devices that are either in production or close to it, while Bosch says it's focusing on the same as it moves forward.

Paul Godsmark, a consultant with Globis Consulting, a company that advises businesses on the impact of autonomous vehicles, and a specialist in intelligent traffic and road safety, thinks that Google may bypass auto makers altogether and sell its technology directly to consumers. He notes that the Washington, D.C., driverless car legislation has a provision that cars less than four years old can be retro-fitted with autonomous vehicle technology.

If Google could get its costs down and sell aftermarket devices to consumers, he says, self-driving cars could be on the road within five years – without input from automakers.

In the meantime, a growing number of U.S. states is warming to the idea. It was only last year that the state of Nevada became the first to license driverless cars, and, already, Google, Audi and Continental have met Nevada's requirement for proof of 10,000 miles of prior autonomous vehicle operation and gained licenses. (For more on the Nevada decision, see Telematics and legal issues with V2V technology.) 

Now California and Florida also have licensing systems for autonomous vehicles. And, according to Stanford University's Center for Internet and Society, 11 other states are considering or have considered bills related to autonomous driving.

ADAS toward autonomy

Automakers and suppliers downplay the autonomous angle, even as they release ever-more-sophisticated safety systems.

Today's cars can already do many things on their own, including park, brake to avoid a forward collision and maintain a safe distance from the car in front.

In March, Mercedes-Benz began advertising its Intelligent Drive system, available on the 2014 E-Class. It includes steering assist, active lane-keeping assist and a brake assist that can detect cross traffic and pedestrians. In mid-May, the company extended the offering to the new S-Class, which goes on sale this fall.

Cadillac's Super Cruise option already approaches highly automated driving: It takes over steering, acceleration and braking, so the driver can relax while keeping an eye on the road.

These systems are all stepping stones to a truly self-driving car.

Sensor fusion

What's needed next is something the industry calls "sensor fusion," which entails combining all the data flowing in through the various sensors and analyzing it to see what's relevant.

Continental has been working on this for three years, and GM used it in the 2013 Cadillac XTS Driver Assistance Package. Ford has said sensor fusion is the next phase in its own intelligent safety systems. As with individual sensors, the processor that handles sensor fusion needs to get smaller and cheaper.

Today's autonomous cars fuse sensor data and analyze it via a laptop tucked into the trunk. To bring these cars into production, OEMs will need to shrink that central processor while making it smarter still.

In Google's case, much of the processing is done in the cloud. The problem with this solution is the risk of dropped cloud connections in areas with poor wireless coverage.

Transform and disrupt

Autonomous cars, in combination with advanced vehicle-to-vehicle (V2V) and vehicle-to-infrastructure (V2I) communications, are expected to spike the greatest changes in the industry. But they are still five to 15 years away. (The time frame depends on whom you talk to.) 

Richard Wallace, director, transportation systems analysis, at the Center for Automotive Research, says V2I and V2V communications are 10 to 15 years out in the United States. “The European Union will be a little ahead,” he says. “[Automakers] are doing it preemptively because they don't want the government to even think about a mandate.”

Simply wresting some control of a passenger vehicle away from human beings could significantly reduce traffic-accident injuries and deaths, experts believe. It could offer the disabled and seniors greater independence. And it could improve air quality if the car's automatic systems drove more economically.

Godsmark says that the societal and environmental advantages of driverless cars would be huge – and that the auto industry should be very scared, indeed. "The automotive business model is to make money by selling lots of cars to individuals,” he says. “The autonomous vehicle blows that paradigm away as there are massive benefits to society and to the average individual by moving to autonomous vehicle fleets."

(For more on the V2X, see Ann Arbor and the future of V2V/V2I, part I and Q&A: Building the infrastructure for V2X.)


A report by KPMG and the Center for Automotive Research concurs. "If self-driving vehicles become a reality, the implications would also be profoundly disruptive for almost every stakeholder in the automotive ecosystem,” it writes.

Reducing crashes, easing congestion and, thereby, reducing the need for infrastructure investment, and getting accurate info on travel times are but a few of the good things to come with autonomous vehicles. 

But car makers would not benefit from this upside. With fewer cars being totaled in crashes, they'd sell fewer replacements. Driverless cars could also reduce the need for personal vehicles, Godsmark says.

Today, a family with two teenagers might have four cars parked in the driveway. But they might make do with two or even one self-driving auto, which could take the kids to school and then return to take mom to work or let dad run his errands. In fact, driverless delivery vans could make it more cost-efficient for dad to have groceries delivered, instead of driving to the supermarket.

Taken a step further, a family car could become an income generator instead of a liability, Godsmark says, thanks to entrepreneurs who will create car-sharing services accessed by smartphone apps. "We will use our car to get to work and then release it to the autonomous fleet, pay a fee to the fleet company to coordinate rides, and, at the end of the day, maybe even have them clean it for us," he says.

Godsmark foresees taxi companies providing autonomous vehicles, perhaps with a ride-along staffer to help with luggage and to serve those requesting help, such as seniors and the disabled. "Taxi, car rental, car share, ride share and parking companies will either adapt to this technology, or they will go out of business," he says. "Once you get to autonomous fleets, society transforms. It will be as big as the introduction of the modern motor car."

Preparing for change 

Automakers are not oblivious to these new possibilities. In 2011, Ford took a stake in the car-sharing service ZipCar. Daimler created Car2Go, a car-sharing service that lets subscribers leave the car wherever they want.

To be sure, this is not a zero-sum game, wherein more cars shared means fewer cars sold, Wallace says. For one thing, if cars could really be made crash-free, they might not need to be so highly engineered, so solidly built, so intensely tested and so highly priced.

The cost savings could be passed on to consumers, making car ownership possible for more people. “If cars were cheap enough, I'd want a crossover utility, a pickup and the Honda Fit for standard commuting,” he says. “Will I actually own those outright or go to a sharing service and specify the type I want today?” There could be new business models, as well, Wallace says, although most of them don't involve car companies. 

Susan Kuchinskas is a regular contributor to TU.

For all the latest telematics trends, check out Telematics Detroit 2013 on June 5-6,Content & Apps for Automotive Europe 2013 on June 18-19 in Munich, V2V & V2I for Auto Safety USA 2013 on July 9-10 in Novi, MI, Insurance Telematics USA 2013 on September 4-5 in Chicago,Telematics Russia 2013 in September in Moscow, Telematics LATAM 2013 in September in Sao Paulo, Brazil, Telematics Japan 2013 on October 8-10 in Tokyo and Telematics Munich 2013 on November 11-12.

For exclusive telematics business analysis and insight, check out TU’s reports: Telematics Connectivity Strategies Report 2013The Automotive HMI Report 2013Insurance Telematics Report 2013 and Fleet & Asset Management Report 2012.

Leave a comment

Your email address will not be published. Required fields are marked *