Visit Your Local PBS Station PBS Home PBS Home Programs A-Z TV Schedules Watch Video Donate Shop PBS Search PBS
I, Cringely - The Survival of the Nerdiest with Robert X. Cringely
Search I,Cringely:

The Pulpit
The Pulpit

<< [ The Bandwidth Game ]   |  Diskless in Detroit  |   [ MEMS the Word ] >>

Weekly Column

Diskless in Detroit: Why Your New Car Doesn't Have a Built-in PC

Status: [CLOSED]
By Robert X. Cringely
bob@cringely.com

Last week in Detroit, there was a two-day conference all about Telematics, which is a made-up word meaning the business of putting computers and digital communication into cars. Telematics is supposed to be a hot area of technical development because we've run out of places to put computers outside our cars, so the next logical place to put them is inside. Except it doesn't seem to be working that way, because at last week's conference, most of the participants seemed to be complaining. Progress was slow, business was illusory, customers just weren't coming in the expected numbers. In short, the car industry was complaining that they'd been "dot-com'd," which is apparently a very bad thing. Telematics is a bust this year, we're told, and might not even deserve to have its own name.

Of course, I have a different view of the whole thing. I think telematics never had a chance in the first place because car companies and computer companies don't even live in the same world.

Before I go much further with this line of thinking, it might be good to cover the logical question of why any reader should even care. Readers in general should care about telematics if they are investors in either car companies or computer companies. And readers of this column should care whether they are investors or not because the story gives yet another insight into how technical organizations do and don't work.

My guide to the world of telematics has always been my friend Shoichiro Irimajiri. Iri is a car guy and spent most of his adult life working for Honda, where he ran the motorcycle racing operation, among many important assignments, and was the guy who built Honda's first U.S. assembly plant on farmland in Marysville, Ohio. Iri worked so hard and well for Honda that he had to retire because of sheer exhaustion. When he caught his breath and decided to work again, he didn't want to work for a Honda competitor, so Iri switched industries, becoming the president and later the CEO of Sega, the video game maker. So Iri knows cars and he knows high tech, and what he's trying to do in his post-Sega career is combine the two. One way he does this is by sitting on the board of directors for Delphi, the world's largest maker of auto parts.

Back to telematics. In the dot-com era, all the market research companies that throw S-curves against a wall and tell us what they think we are going to be buying a few years down the road were predicting we'd be buying a lot of cars filled with computers and computer services. Of course, these same folks also predicted the death of television and the end of retail stores, so why should we be surprised that they were wrong about telematics, too? But why didn't it work? Why aren't our cars filled with personal computers and why aren't we web surfing our way down the road? It all comes down to two simple facts, both of which were obvious from the beginning to anyone who would bother to drop their S-curve for a moment and put on their thinking cap. Telematics hasn't worked so far because we don't have hard disk drives that will reliably run in automobiles, and telematics hasn't worked so far because it takes four years to design a car.

Computers in cars require storage, lots of storage, especially since they tend to be disconnected from the rest of the world and yet still have the need to carry maps of every back road between here and Possum Trot, Arkansas. But cars present a hostile environment for computers. They have to operate in freezing cold and blazing heat. Leave a CD on the black dashboard of your parked Chevy in the summertime, and you'll know what I mean. And then there's vibration. Cars drive through potholes, and disk drives hate potholes. So there simply isn't a disk drive today that is built in volume at a good price that can live for 100,000 miles in nearly any car.

But why aren't there any cheap hardened disks? Cars would seem to be an enormous market — nearly as big as personal computers. Why can't the hard disk makers get their act together to build car drives? That's where the four-year car development cycle comes in. Cars take four years to design, which means that right now Detroit (and Tokyo and Stuttgart) are hard at work on the specs for the 2006 Wombat GT. To be included in those specs, a hard disk maker like Seagate has to already have in hand the design for its 2006 disk drive, which of course, they don't.

The hard disk business is so terrible that most makers today can't guarantee they'll even be in business in 2006, so they won't devote a penny to planning products that far in advance. Hard disks these days follow a six month product cycle and most companies look two cycles ahead, which means 12 months. That's the horizon on the hard disk world. Looking beyond that horizon is just too risky because technology changes, as do market needs. Who knows what personal computers will look like in 2006? I don't, Bill Gates doesn't, and the product planners at Maxtor don't. We don't even know absolutely for sure that there will even BE personal computers in 2006.

Of course, hard disk makers could build drives that meet the 2006 needs of Detroit, but they'd need Detroit to start buying those drives a year from now, something Detroit won't do. And so telematics suffers.

Is there a way around this problem? I think there is, but it requires some imagination on the part of the car companies, which seem to be enjoying all too much their current role as high-tech whiners.

There are only two real hopes for telematics. First is the aftermarket industry. If we ever do get a hard disk drive in our cars, the first one will probably be installed at a car stereo shop, not at the factory. Alas, that isn't a big enough market to entice the beleaguered hard disk makers, who can't afford to lose even more money designing for what would initially be a very small volume business. The second hope for telematics, though, is that the car makers could bring themselves to think literally outside of the box. Instead of seeing the car as something to put intelligence in, they could see it as a platform to put intelligence ON — intelligence that might have nothing at all to do with transportation.

Cars are the perfect Trojan horse for distributed communications, for example.

Cars are everywhere people are, they are generally outside, they have their own power source, and they have extra places to stash black boxes.

A really smart car company might take a look at Mesh Networks, for example. Mesh is a Florida company that is about to introduce a proprietary 2.4 GHz wireless network that offers dramatic advantages over 802.11 WiFi. Mesh nodes act as routers and repeaters so communication can be extended far beyond normal Internet access points. Mesh networks support tens of thousands of router/repeaters.

Mesh networks offer Quality of Service (802.11 doesn't) and support Voice-over-IP. An enlightened car company — or better still EVERY car company — should put a Mesh node in every car they make whether the owner wants it or not. In a couple of years, when 20 million Mesh'd cars are on the road and the car companies (I'd suggest a consortium of car companies, since this is a new field for them and anti-trust is not a concern) could light that network and, in one stroke, take a big chunk of the U.S. telephone, Internet, and mobile phone markets. Just buy space on cellphone towers and tie it all together with cheap fiber from a Global Crossing or Williams Communications.

Of course, it isn't really that simple. My ideas never are. But it is simple enough that Toyota is considering doing exactly this in Japan. Watch out NTT DoCoMo.

The network I describe could use better receiver technology, for sure, but that's available from another company called Wiman (check the Links of the Week for this and other interesting info). But what's the risk to these enormous companies, really? It has to be better than zero percent car financing. Throw an extra $20 to $30 into each car for a chance at $100 billion two years from now. That's half the time it takes to design a car, but it's okay because this service is purely parasitic, just bolted on. And it doesn't even require a hard disk drive.

Comments from the Tribe

Status: [CLOSED] read all comments (0)