Don't Drive Ahead on AutomaticBy Tom Steinert-Threlkeld | Posted 2002-11-01 Email Print
WEBINAR: Available On-Demand
Innovate and Thrive: How to Compete in the API Economy REGISTER >
Even if fail-safe software arrives, pleasing customers will still take a smart staff.
Pretty soon, you will be driving an information system.
You might think you're already there. Your car today may have 50 microprocessors hidden inside it, helping regulate everything from its speed to the sound coming into your ears.
But these are a collection of controllers; they are not a system. In a decade or so, a real system of electronics is likely to replace the engine, pedals and steering wheel in your car.
Instead of that wheel, you may find a pair of handgrips. Instead of an engine, a stack of fuel cells and an electric motor will supply powerbut you won't even see them. The entire "drive-by-wire" system will be enclosed in a thin slab of controls called a skateboard, under the passenger cabin. A chassis will be a thing of the past. When you want a new look, you'll simply replace the shell.
The future doesn't always arrive, of course. Otherwise, we'd all be sitting back on automated highways by now. So don't be surprised if the "hy-wire" approach now being propounded by General Motors, the subject of our November Case Dissection, doesn't turn out quite as envisioned.
But the broad implications of the digitization of the automobile should interest you, as technology managers and business decision-makers. Where GM is headed brings into high relief the kinds of reliability, customer service and privacy issues a wide range of industriesfrom airplanes to appliancesare likely to face as they build increasingly complex information systems into their products.
Take a look at what GM puts into the 2002 versions of most of its cars. Its OnStar satellite information system now serves 2 million vehicles and is adding 4,500 customers a day, according to GM's chief technology officer Tony Scott.
That translates every month into 200,000 calls for route information, 14,250 for roadside assistance and 15,000 for the remote unlocking of doors.
Plus, the system helps recover 375 stolen vehicles every month. Think about that.
OnStar keeps track of where any subscribing vehicle is located. It's not much to ask the electronics in a system such as this to turn off the motor, when driven by an "unauthorized user."
What sheriff, police chief or federal marshal anywhere in the country wouldn't leap at the chance right now to stall every car for 20 miles around, after another sniper shooting?
That's scary. It's kind of like putting a cryptographic Clipper Chip in every vehicle, to serve and protect. Are we ready for this?
Not yet. Maybe not ever. That strikes at the core of the ultimate American value: freedom. To its credit, GM today does not run a vehicle location service, except in cases of theft. It has decided not to abet individuals who simply want, for good or bad, to track the whereabouts of spouses or teen-agers.
These are the kind of privacy and customer relations issues manufacturers will face, as the software components of their products escalate. (See our related news story.)
But the issues won't be just philosophical. For GM, it gets down to the reliability of its products. If GM can't manage the transition to the "drive-by-wire'' car effectively, it will run itself out of business.
Simply put, software right now is not good enough. Microsoft may relentlessly improve its products, but, as GM's Scott repeatedly pounds home, it "just won't be an option for us" to tell customers "to hit control-alt-delete to reboot the car."
In transportation, fail-safe software will be a must, as will around-the-clock automated and manned customer support. Whether it's in a data center in Warren, Mich., or a service bay in Altoona, Pa., car mechanics now are also going to have to be systems engineers.
The same will apply to Sears, Best Buy and other retailers trying to win and keep customers in the future. The back shop will take center stage; and those that don't focus on supporting customers of data-driven products after the purchase will fail, as surely as those who don't pay heed before the purchase.