GM's wireless data service is about to turn your car into a four-wheeled vertical portal.
If you haven't kept your eye on the horizon, you might think General Motors is only involved in the old-fashioned business of making and selling automobiles. Cars are still what the company does best, of course: One of its vehicle lines alone - the GMT 800, a full-size truck sold as the Chevrolet Silverado or GMC Sierra - will generate roughly $18 billion in revenues this year, making it a bigger enterprise than Microsoft.
But for GM, the future is just as much about serving the driver as it is about selling the ride. The 500 million hours a week that Americans spend in their cars represents a huge audience of consumers. And that explains why capturing subscribers, not harnessing horses, is the hot new motor trend, as an emerging constellation of automakers, electronics manufacturers, and telecom providers convert transportation platforms into communications platforms that connect the driver, and the vehicle itself, to the rest of the datasphere.
The name of this game is "telematics" - a buzzword coined for the convergence of wireless technologies, global positioning systems, and onboard electronics. The auto giants have long talked up a future in which smart vehicles zip along smart highways, and GPS navigation units have already become the nerd equivalent of a CD changer. Telematics connects the dots between the speculative hype and the new infotainment age by putting at the driver's fingertips a bundle of safety, route guidance, convenience, and entertainment services. Within a few years, motorists will take for granted reminders to head south at the next interchange, or news that a Chevron station and a Burger King can be found up ahead. Before long, your car may even email you an alert that its transmission will soon require maintenance.
Only 80,000 cars in the US are equipped with telematics hardware now, according to Stephan Beckert, senior consultant at the Strategis Group in Washington, DC. But by 2003, he predicts, that number will climb to 1.2 million, with hardware sales of $240 million and another $162 million in service revenue. Demand could really take off when today's blue-sky experiments in, say, backseat satellite TV start to hit the showroom. "The market is still in its very early stages," Beckert says. "To use a PC analogy, we haven't even reached the Commodore 64 plateau yet."
Telematics systems, unsurprisingly, made their debut inside luxury models produced by the likes of Lincoln, Cadillac, and Mercedes-Benz. But the vast majority now on US roads - some 75 percent - were built by General Motors. "For now," Beckert says, "GM defines the entire telematics market. No one else has picked up the ball."
General Motors owes its lead to a small service division called OnStar. With just 250 employees, OnStar is a speck on GM's corporate org chart, and its subscriber base of some 55,000 customers isn't much bigger than that of many mom-and-pop ISPs. Inside the car, OnStar's interface is equally unassuming - a hands-free cellular phone that links to a call center via three buttons on an overhead console just above the driver's right ear. Today OnStar is standard only on the Escalade, Cadillac's new SUV; otherwise it's a dealer-installed option, typically selling for $700.
Yet by 2002, GM intends to make OnStar - now available on more than one-third of its vehicle models - standard equipment on every one of the 5 million cars and trucks that roll off its assembly lines each year. Multiply that by the average $25 a month the service collects from subscribers, and OnStar could quickly deliver annual revenues of some $1.5 billion - and a huge pool of content consumers.
"We start with a carrying platform called cars, which is really an opportunity to grab a captive subscriber base," says Ron Zarrella, president of GM North America. "From there, we can leverage that relationship to provide more and more telecommunications and telecomputing services."
General Motors has a big head start, but the competition is closing fast. The new Mercedes S-Class, for instance, comes with a three-button service called Tele Aid in addition to a spiffy in-dash LCD navigation display. And automakers aren't the only ones trying to plug in to the 15.6 million new vehicles sold in the US each year. America Online, a company famous for having grown its proprietary data service into a booming content platform, is angling to serve its users via wireless devices. Meanwhile, Microsoft has already established a dashboard foothold in the form of the Clarion AutoPC, a WinCE-powered cross between a car stereo and a set-top box.
Competition aside, some major technical obstacles stand in the way of OnStar's ambitions - lowering the cost of telematics hardware, standardizing cars' electrical architecture, and cloaking the nation in wireless bandwidth. Yet if everything works out as GM has planned, OnStar will be the biggest portal on the rolling media platform the automobile is destined to become.
With her computer monitor flashing red alert, 21-year-old Marina Ivezaj is anxiously responding to a collision in Largo, Florida, that triggered an air bag deployment in a 1997 Cadillac DeVille. OnStar automatically notifies a 24-hour call center in suburban Detroit whenever an air bag goes off; the car's position appears on Ivezaj's monitor as a dot on a digitized map, while a popup box gives the number for the nearest 911 center.
"This is OnStar," Ivezaj says into her headset. "We've received notification that your air bag has gone off. Are you OK? Do you need medical assistance?"
There's no answer. She can hear people inside the car - "they're freaking out," she says - and they presumably can hear her via OnStar's audio interface. Meanwhile, she's on another line, telling the 911 dispatcher to send out an ambulance. It's a tense situation - a verité version of OnStar's commercials touting the system's dramatic ability to pluck motorists from the path of peril.
Still no response from the Cadillac. But off in the distance, Ivezaj can hear an approaching siren, growing in intensity until it almost sounds like it's inside the car. An emergency team is on the scene; her mission is accomplished.
With more than 60,000 service requests pouring into OnStar's two Detroit-area call centers each month, minidramas like this unfold almost daily. There are ambulances sent to the scene of roadside heart attacks, tow trucks directed to help stranded motorists, and police put on the GPS-tracking trail of stolen vehicles. If you lock your keys in the car, an OnStar adviser can open the doors remotely by clicking an onscreen icon. If a service light begins to shine, an adviser can diagnose the problem by sending a query to the onboard engine-control computer.
In practice, advisers function as geographic clairvoyants, reference librarians, and wireless magicians. The call center staff consists mostly of young college grads like Marina Ivezaj who undergo a 60-day training program that emphasizes map-reading, familiarity with OnStar services and how they're delivered, and good old-fashioned people skills. Backstage, the system relies on a GTE Wireless Win4 data network and OnStar's proprietary database, which includes roughly 5 million location entries searchable by name, business category, and proximity, to help motorists find anything from bank machines and racetracks to government offices or the nearest karaoke bar.
OnStar's original interface, based on a programmable cell phone that required a separate service contract, is being phased out this year in favor of the "Gen 2" all-in-one design. Basic service runs $199 a year for the lifeline package of safety and security; the $399 premium bundle tacks on unlimited route support and locator assistance, plus OnStar Concierge - a service that can help you order flowers or track down Super Bowl tickets.
Behind the wheel, OnStar is designed to keep driver distractions to a bare minimum. The GPS receiver and 9-inch whip antenna sit discreetly on the car's exterior; the overhead three-button console (linked to an embedded 3-watt phone) is inconspicuous but close at hand. To start a typical OnStar service request, you press the blue button. A recorded voice - sounding a bit like HAL's spooky kid sister - chimes in over the car stereo: "Connecting to OnStar." Then a cheerful adviser greets you by name, identifies your location, and asks how he or she may be of service. In more grave situations, the red emergency button gives priority access to the call center. And, of course, if the air bags inflate, the call center is alerted automatically.
The service has struck a chord. According to research by J. D. Power and Associates, 42 percent of Escalade buyers said OnStar was a factor in their decision to purchase the vehicle, while 23 percent said it was the reason they decided to buy a Cadillac over a competing luxury SUV like the Lincoln Navigator. As of this spring OnStar is an option across much of GM's product line - from the dowdy Chevrolet Lumina to soccer-mom minivans like the Oldsmobile Silhouette. Though the company's not yet equipped to make much of the marketing data it obtains, call center requests suggest middle-of-the-road tastes like Red Roof Inns and McDonald's. Overall, the service has major mass-market potential.
But perhaps the biggest asset is OnStar's friendly, simple interface. With call center advisers serving as the primary link to people in their cars, OnStar has generated an intense brand loyalty and sense of community that would be the envy of any wannabe ecommerce portal. In focus groups, users cite Mother Teresa and Colin Powell as the public figures who best personify the system, according to OnStar marketing manager Rodney Williams. "Cadillac owners are hardly what you'd call early adopters," Williams says. "But their reaction to this technology has been amazing."
And there's plenty of blue sky ahead on the technical front. "We know that safety, security, and peace of mind anchor this market," says Chet Huber, OnStar's managing director. "Beyond that, we come to convenience and entertainment, while creating additional opportunities to layer value in the relationship. We've papered the walls with ideas. We could give the best package tours in the country. If you see a sign for the Gettysburg exit, we could offer you the opportunity to download a cool tour by tapping into the GPS and your stereo system. As you drive by Little Round Top, you'd get a history lesson, and as you pass Pickett's Charge, you'd hear cannonades."
For now, OnStar is experimenting with systems that can deliver audio Internet content such as stock quotes, sports scores, or weather reports. There's talk of remote vehicle scans transmitting engine performance information to a personalized Web page that can predict future maintenance requirements. Productivity services, based on technologies like Bluetooth or Firewire, aim to connect dash-mounted devices like Palms to networked PCs on distant desktops. Further out, drivers might alter engine or suspension performance by calling up OnStar and ordering algorithms that can be downloaded to the vehicle.
Founded in 1908 as a collection of free-standing automakers under a single corporate parent, General Motors emerged as a symbol of US industrial might under the leadership of Alfred P. Sloan Jr., GM president or chair throughout the boom years from the 1920s to the 1950s. "The rapidity of modern technological change makes the search for facts a permanently necessary feature of the industry," Sloan wrote in his 1963 autobiography. "That seems obvious, but some of the biggest changes of position in the industry came about in part because someone got an idea he thought was eternal."
The failure to heed Sloan's call for constant innovation came back to haunt all of Detroit's automakers after competition from Europe and Japan began arriving in the 1970s. And while Chrysler was famously pushed to the brink of bankruptcy in 1981, over the long haul GM has come out the big loser. In the last two decades, both Ford and Chrysler have managed to increase their overall market share - from 20 to 25 percent in the case of Ford, and from 9 to 16 percent at Chrysler. GM, meanwhile, has watched its portion plunge from 45 percent to its current 29.
Even so, General Motors remains the biggest company in the world, with 1998 revenues of $161 billion and 594,000 employees scattered around the globe. And with that huge size comes diversity. Apart from GM's core automotive brands - Saturn, Chevrolet, Pontiac, Oldsmobile, Buick, Cadillac, and GMC - the company's reach extends from GMAC home financing and auto insurance to Hughes satellite telecommunications and, by extension, into 5 million American living rooms through GM's other hot media property, DirecTV.
OnStar's origins can be traced to late 1994, when a group of engineers from Electronic Data Systems - then owned by GM - conceived a plan to offer onboard data services via microwave transmission. "The telecommunications and computer industries had started with the office and gone to the home - the logical next step was the motor vehicle," recalls GM vice chair Harry Pearce, who at the time oversaw both EDS and Hughes Electronics. "But the car was conspicuously disconnected."
Having once earned his living as a hard-nosed trial lawyer, Pearce is known at GM for his aggressive efforts to use the company's automotive expertise to create new businesses. "My vision from the get-go was that this technology was not simply another vehicle feature," he says. "It was a unique opportunity to get into the telecommunications services business, where you get downstream revenue, better margins, and, most significantly, great growth potential for General Motors."
Undaunted by the huge capital costs of building the microwave platform, Pearce recruited a few Hughes engineers to examine the idea; they instead suggested using the existing US cell phone network. And with that change, the proposal - now code-named Project Beacon - suddenly became financially viable.
In 1995 Pearce tapped Chet Huber to bring the Beacon project to market, under a new General Motors division called OnStar. Having spent 23 years selling GM locomotives, Huber was fueled by a salesman's scrappy enthusiasm, and he attacked his new task with the fervor of a Linux developer. "The car deserves a dial tone!" Huber declares. "It deserves to be a node on the network! It wants to be a part of the infrastructure!"
In the spring of '96, just a few months before OnStar made its commercial debut, GM gave a presentation at TED, the exclusive tech confab held each year in Monterey, California. Ron Zarrella - then VP of sales, service, and marketing - and Vince Barabba, GM's general manager of corporate strategy and knowledge development, took the stage by modestly acknowledging their roots in "the anachronistic rust belt called the Midwest." Then, over the next hour, they outlined how OnStar embodied GM's broader shift from a company that mechanically built "a car for every purse and purpose" (as Sloan had decreed in the 1920s) into a more nimble company that would produce "a car for every purpose, functionality, need, want, attitude - you name it."
During the Q&A session that followed, Adobe CEO John Warnock asked whether OnStar would be an open or closed system. Caught off guard, Zarrella and Barabba admitted they had not thought that through yet. Afterward, however, they started asking around, and their inquiry crystallized a reassessment of what OnStar would become.
The question of open versus closed represented a fundamental challenge to GM's business model. Ever since Henry Ford began building Model Ts in "any color you want, as long as it is black," the auto industry has treated the vehicle as part of a proprietary OS that extends to every aspect of product development. Beyond such basics as gas, tires, and spark plugs - and aftermarket gadgets like stereos and alarms - the car is essentially a closed platform.
The digital world, of course, is increasingly "platform agnostic" - value rises as the number of users and product developers increases. For GM, the potential benefit of this type of thinking soon became obvious. By turning the system into an open platform, OnStar would be able to harness the entrepreneurial energies of thousands of independent hardware, software, and content developers.
"We finally realized we don't have all the ideas," Barabba says. "Once we get people thinking about the car as an information platform, then ideas that we never even dreamed about will start to surface."
OnStar's business offices are in a tall tower in Troy, Michigan, 30 minutes up I-75 from GM headquarters in downtown Detroit. The tower itself is an anonymous structure that could just as easily be perched along Highway 101 or Route 128, although the surrounding parking lot betrays one unique local characteristic: virtually all the cars here are American-made.
Inside, it's a pretty typical working arrangement, with a labyrinth of beige cubicles in the center and a ring of private offices lining the perimeter - just another hive of casual young professionals buzzing with the energy of a fledgling tech startup. That's largely because more than half of OnStar's employees came to Troy from firms like FedEx, Johnson & Johnson, GTE, Rockwell, Ameritech, and Bell Atlantic, and they're all looking beyond Detroit to capitalize on OnStar's early head start.
OnStar's future now sits in the hands of chief vehicle engineer Dave Acton, an auto industry veteran who worked the GM assembly line straight out of high school but knows his future lies in the network economy. A graduate of the General Motors Institute (now Kettering University), with an MBA from the University of Michigan, Acton rose to director of electrical engineering at Cadillac. In 1997 he was promoted to a similar post for all of GM North American Operations.
Since coming to OnStar in 1998, Acton has been tackling one of telematics' most nagging problems - the auto industry's long product cycles. "Even if GM gets its vehicle development cycles down to two years or 18 months, that still represents about three generations for most consumer electronics products," he explains. "When the first car rolls off the assembly line, the onboard electronics will already be obsolete. The trick is to install current-generation electronics on new cars before they roll out the door, while also allowing for upgrade capability over the course of the vehicle's useful life."
That life has already been extended by the redesign of automotive electrical systems into a kind of onboard LAN. Unlike the old tangle of copper wires and harnesses, today's architecture is organized into a series of microprocessor-based control modules, each governing a specific set of vehicle functions. Thus a car door contains a module that governs the use of accessories like power windows, door locks, and mirrors, while modules elsewhere control antilock brakes, air bags, audio components, and engine performance. There are upwards of 30 modules in contemporary luxury cars, all linked together via a shared data pipeline that runs throughout the vehicle.
While the switch to LAN-based architectures has dramatically reduced the number of wires - making cars both easier to build and more trouble-free over the long haul - it's also forced a closer linkage to each carmaker's unique specifications. And that, in turn, has created a knotty set of challenges for any consumer electronics supplier who dreams of selling accessories like audio equipment, antitheft alarms, navigation units, or seatback video displays.
To illustrate the problem, Acton likes to tell a story about testing out a prototype '98 Cadillac Seville STS. A group of senior GM execs was on hand to watch as the new vehicle was put through its paces, but when the key was turned, the Cadillac wouldn't start. The fault was traced to a stereo amplifier built by the Bose Corporation; a subtle design flaw had sent corrupt data coursing through the car's network, effectively crippling the starting system. "It was serendipitous," Acton says, "because it allowed me to demonstrate just how important all this integration stuff really is."
After getting GM's management to sign off on a uniform "gateway" that would shield critical vehicle systems yet make it easy for third-party suppliers to plug in, Acton started networking with other auto engineers. In May 1998 the initial cadre of gearheads was expanded, and by October the group had become the Automotive Multimedia Interface Consortium, or AMIC - a collaboration representing Chrysler, Daimler-Benz, Ford, General Motors, Renault, and Toyota. The vision: to develop common standards for jacking aftermarket information and entertainment devices into the central nervous system of any car or truck. Taken together, AMIC's members now build more than half of the 55 million cars sold worldwide each year.
On a nuts-and-bolts level, AMIC will establish specifications for four key interface characteristics. The first is a physical interface for connecting a device to a car - the "plug" of plug-and-play. The second is a virtual interface, or software gateway, for safely transferring data between aftermarket devices and each carmakers' in-vehicle LAN. Third is a uniform set of data protocols - specifically, a low-bandwidth interface (for use with simple devices like pagers, telephones, and infrared tollbooth payment systems) and a high-bandwidth interface (for large data transfers, high-definition video, and CD-quality audio). Finally, AMIC will likely use Java as an application programming interface, so coders can easily write software that will run on any vehicle platform.
Practically speaking, the AMIC specs will allow car electronics to expand at the same pace as consumer electronics. For example, if you buy a car in 2002, when the first AMIC-enabled vehicles should begin to arrive in showrooms, you won't miss out if Star Trek-style voice systems don't debut until 2006 - the upgrade can simply be plugged in to your car.
The goal, in other words, is a "write once, run anywhere" environment for cars - an open platform that will make it easy for aftermarket suppliers to access the data-moving capabilities of telematics systems like OnStar. So far the effort has been warmly received. "Developing a standardized interface will be a huge step forward," says Bob Locascio, executive director of the Telematics Suppliers Consortium, which represents vendor companies such as Motorola, Hewlett-Packard, Kenwood, and Alpine Electronics. "We don't care too much about the details - we'll love it, so long as manufacturers resist the temptation to distort the standard by developing their own unique flavors."
AMIC represents a decisive step toward turning the automobile into an open platform for onboard hardware, software, and wireless services. With the addition of AMIC, OnStar will be able to offer any third-party content provider or electronics manufacturer access to OnStar subscribers - in exchange for a share of service revenue they'll generate if their applications find an audience.
For the moment, however, those applications are hamstrung by the analog cellular phone that serves as OnStar's primary conduit to the outside world. Analog cellular delivers pervasive, nationwide coverage, but at a price: low, low 2400-baud bandwidth that can't support simultaneous voice and data communications. The trickle forces OnStar's advisers to put customers on hold while remotely updating a car's GPS position or unlocking its doors. It's impractical to even contemplate downloading detailed map images to the car, to say nothing of on-demand video.
Eliminating this bottleneck won't be easy, and there's a broader push afoot among telecom providers to build a common platform for all wireless applications. (See "Air Apparent," page 133.) Unlike in Europe, where the GSM digital wireless standard prevails, the bandwidth challenge is compounded in the US by two competing standards. Cost reduces the likelihood that cars will hook into global satellite webs like Iridium or Teledesic anytime soon. However, "the cost of delivering service can be significantly lowered by broadcast-based technologies," according to Brian Gratch, director of marketing for the Telematics Communications Group at Motorola, which supplies the hardware for OnStar and most other systems. "There's some really cool stuff we can do using satellite radio, or FM subcarrier frequencies, or HDTV sideband." (In the meantime, competing telematics services are reaching drivers in a variety of forms - see "The Telematics Options," page 129.)
Another one of OnStar's nagging shortcomings - the system's inability to give customers a visual representation of their current location and destination - can be eliminated by the addition of an LCD screen and an onboard CD-ROM map library. But such nav systems have a downside: Trunk-mounted CD-ROM libraries require regular - and expensive - upgrades. Even worse, reading a map displayed on a center-mounted console is only slightly less hazardous than driving with a road atlas on your lap.
Yet even audio-based turn-by-turn navigation won't eliminate safety concerns. "Voice interfaces are not a panacea," says a National Highway Traffic Safety Administration official who has studied in-car computing. "Our evidence indicates that people have a tendency to look at any device, such as a visor-mounted microphone, if they're talking to it. Basically, any requirements you add to the driving task - whether cognitive, visual, or motor skills-oriented - ultimately increase the accident risk."
While carmakers are gearing up for the telematics race, some big names from the computing world have already started to break into the market. This year saw the arrival of the WinCE-based Clarion AutoPC, a dashboard device that's about the size of a car stereo. Using computer-generated speech and voice-recognition technology, AutoPC responds to your commands and serves up optional services like concierge assistance, paging, email, and real-time traffic reports.
Microsoft, unsurprisingly, aims to have some say in defining the in-car data platform. "Just as PC manufacturers can choose different operating systems and applications, maybe car manufacturers will choose an OS based on the quality of the product," says Phil Holden, Microsoft's group product manager for Windows CE. "For now, we're working very closely with the auto industry, and with the AMIC folks, to move things forward. We see ourselves supporting open standards when appropriate, and then building value-added products on top of that."
The big Redmond machine has already altered the direction of in-car communications. "Microsoft was definitely a catalyst for the creation of AMIC," says Paul Hansen, who writes the respected technology newsletter The Hansen Report on Automotive Electronics. "Carmakers - especially the Japanese - were very concerned that Microsoft could try to define an in-car OS that would then become a standard. They didn't want to let that happen."
"Automakers have seen what Microsoft did with desktop operating systems, and they don't want to see that in their industry," adds James DeStefano, Sun Microsystems' strategic marketing development manager for automotive applications. "Windows will be just another application running on the car - and it will probably be a very good application for mobile office functions."
But productivity apps are only part of the telematics picture. America Online is also moving to find an in-car beachhead. "We have a plan for all alternate devices - anything that isn't a computer," says Barry Schuler, president of AOL's interactive service group. "Our strategy is that when these devices come out, we'll support them with basic services for our members, then make decisions when the devices hit critical mass as to whether we're going to build full-fledged services. We're pretty comfortable that we'll be the company people want to partner with."
For now, OnStar's edge continues to be its intimate connection to the biggest car manufacturer in the world. Yet that same connection may one day pose its own problems. "GM has a very significant challenge in terms of fitting its business plan into the overall structure of the computer and communications industry," argues Sun's DeStefano. "Consumers are already used to the channels that they currently have in their home, office, and portable applications, and they're going to expect to have many of those same channels available. Do you really want to have one set of destination services if you drive a BMW, and another set of services if you also own a GM car? Customers will want continuity, and how that continuity works out is going to be an interesting issue in the evolution of the marketplace."
As the automobile is steadily transformed into a fuel-injected content platform, new brand names will no doubt emerge - Amazon-style - in the quest to pump personalized services and information into the cockpit. Until then, OnStar is likely to remain the biggest player on the telematics frontier. "It's GM's market to lose, because they are defining all the rules," says Michael Kujawa, a senior analyst with Allied Business Intelligence. "OnStar has the potential to become another Kleenex - a generic brand name for all telematics products."
Yet easily obscured by the promise accompanying this grand convergence of automotive and information industries is the more practical issue of what drivers really want as they watch the world go by through the windows of their speeding automobiles. Rather than simply reinventing the dashboard for the digital age, OnStar redefines the man-machine interface, using a helpful voice on the other end of a hands-free phone - it's a relationship with technology that's as comforting and familiar as speech itself.
And in the end, that might be GM's greatest contribution to the four-wheeled media platform - the realization that OnStar's future as an in-car portal will be secured by the human touch. Because when you get right down to it, that's what telematics is really all about. We may roam, and we may wander, but we'll never really be alone in our cars again.
PLUS
