I recently had a conversation with Dieter Zetsche, the Daimler CEO. He had arrived at a social event in Frankfurt in the back seat of a driverless car. It was the new S-Class Mercedes in a demonstration of autonomous driving.
I said to him, “You can load all this self-driving technology into an S-Class because they sell for a $100,000-plus, but will this stuff ever reach the lower-priced cars?” He looked at me as though I had a screw loose and said, “Yes, easily. It’s only code.” Pause. “You know, software.”
Writing code is a lot cheaper than bending metal. Zetsche is a man accustomed to spending huge amounts of money. The price tag to develop a new vehicle starts around $1-billion and can run to five or six times for an all-new platform with new powertrains. You can have rooms full of software developers in their propeller beanies and not notice the cost in a game like this.
Big Data is a big deal for next-generation cars.
There are already hundreds of sensors in a new car connected to microcontrollers that run the engine, brakes, airbags, traction control, the works. It’s all “drive by wire” these days. Pressing the accelerator no longer pulls a throttle cable; the steering wheel is no longer directly connected to the wheels. Sensors and data processing are making most of the decisions, and adding a little more “code” will deliver many more functions at little extra cost.
Gyroscopes and accelerometers are already in place for anti-rollover systems. GPS is practically standard these days. Ultrasonic systems and radar and stereo cameras are already watching traffic, following lane markers, reading signs and detecting accidents before they happen. The next thing coming is vehicles that communicate with the driving infrastructure (V2I) and vehicles that communicate with other vehicles (V2V).
The assumption is, if you can predict the accident, you can prevent the accident. Well, you can’t, but the technology can and it will make the car drive itself out of the way.
Getting vehicles to talk amongst themselves and communicate with the “cloud” means that whoever is in the vehicle is telling the world a great deal about themselves. Like, why were you at this sketchy location at 2 a.m. and why were you driving so badly? And why do you always buy your jug of milk (or quart of Scotch) at that store when it’s cheaper on the next block with the coupon that was just sent to your car?
A great example how data can be used against you was the way Tesla hung out to dry a car reviewer for The New York Times. He had claimed in print that the all-electric Tesla had run out of juice early and had to be towed home.
However, Tesla lashed back by releasing the raw data from the vehicle that showed that charging stations were driven past, that the car was unplugged before it reached full charge, that it was driven at greater speeds than claimed and even that the heater was on high. In other words, the data shot his story and his credibility full of holes.
To make the most out of Big Data, vehicles will have to communicate a lot of information about you, whether you like it or not. If you have a plug-in vehicle, the electrical utility will use artificial intelligence programs to manage where and when you plug in, so everyone doesn’t plug in at once and crash the aging grid. And, when they know your driving habits, they’ll know you don’t need 30 amps and will only dish out 10 if there’s high demand.
Big Data can only deliver big benefits with data-sharing cars. Will it also be a big infringement on our privacy? It sure looks like it. But maybe it’s worth it if we get better fuel economy, get stuck less in traffic and actually have way fewer accidents. It’s possible and it costs little. “It’s only code.”