GM’s Internet Overhaul
The Advanced Design Studio on General Motors’ massive technology campus in Warren, MI, feels like a Silicon Valley software startup from the halcyon days. Three six-meter “power walls” snake across one section of the room, displaying far-larger-than-life, 3-D projections of vehicles in progress, for everyone to examine and dissect.
In a dimly lit corner of the room, engineers, designers, digital sculptors and software programmers sit side by side, their faces illuminated only by the light from their computer screens. Alan Rhodes, vehicle model manager of the studio, addresses the Smart Board, a 127-centimeter flat-panel computer display synchronized with the power wall. With the tip of his index finger, he maneuvers 3-D sketches of upcoming GM vehicles around the screen, alternating between views of the Chevy SSR (Super Sport Roadster)-a retro-styled convertible pickup truck unlike anything GM has ever built-the Pontiac Solstice concept car and a new Hummer called the H2.
Designing and displaying vehicle models in software is a relatively new capability for GM, the world’s largest manufacturer of cars and trucks. Just a few years ago, all sketches were created by hand and stapled to a display board for review. Clay or hard Styrofoam models had to be sculpted for each new design.
Now, with all 14 GM engineering centers digitally synchronized via a corporate intranet and all using standard computer-aided-design software and 3-D simulation tools, designers in Holden, Australia, Russelsheim, Germany, and North Hollywood, CA, can collaborate around the clock. Most important, the new technology allows GM’s senior management to review fully realized designs in a far timelier manner than ever before.
“We’re using the technology they used to make Toy Story,” Rhodes says, referring to the hit film digitally animated by Pixar Studios. “With this, we can get feedback right away. We integrate real people into the animation so that we can check spatial relationships within the cockpit of the vehicle.”
Indeed, Rhodes points out that several recent hires have come from the motion picture industry. “We used to hire strictly fine-arts majors and automotive-design grads,” he says. “Now we’re recruiting computer game designers and people from the film industry.” To get his team’s creative juices flowing, Rhodes sometimes runs films such as The Matrix on the power wall during lunch hours.
None of the cars whose high-tech images flashed on the power wall resemble your father’s Oldsmobile-or Chevy, Buick or Pontiac. But this is not your father’s GM. The Advanced Design Studio is part of one of the largest technological overhauls in corporate history. All told, the $180 billion automotive behemoth has invested more than $1.7 billion in Internet applications while eliminating 3,500 older, or legacy, information systems. By renovating its bloated and outdated technology infrastructure, GM has become nearly unrecognizable as its former self and, without much fanfare, changed nearly a century’s worth of automaker tradition in Detroit-and, possibly, the world.
Vehicles that once took four years to design and build are now being created in 18 to 21 months, as GM pushes for greater efficiencies and quality. The company is now introducing a new vehicle every 23 days. That translates to twice as many annual product launches as at any other period in its history. The cumbersome and expensive physical processes of model building and crash testing have been streamlined by the introduction of digital simulation technology. The technological changes even seem to have unlocked designers’ creativity, once the victim of GM’s overwhelming bureaucracy, resulting in a generation of cars and trucks with radical designs.
Such rivals as Ford, Daimler/Chrysler and even Toyota, which was long renowned for its homegrown and efficient product design process, have made similar commitments to technology. But GM is clearly leading the way, according to Thilo Koslowski, lead automotive analyst for the research firm Gartner. “Even compared to the Japanese,” he says, “GM has been the most aggressive company in using technology to change business processes.”
Remaking a Giant
It would be difficult to find a corporate giant more unlikely to achieve information age glory than GM. Tagged for much of the past three decades with adjectives like staid, slow-moving, bureaucratic and moribund, GM had been slowly suffocating from its own inertia. Plagued by bloated cost structures, labor troubles, forgettable products and intensifying global competition, GM had watched helplessly as its market share slid from nearly 60 percent in the 1960s to less than 30 percent by the 1990s. As the millennium approached, GM seemed stuck in the slow lane as its big-three rivals Daimler/Chrysler and Ford Motor prepared to roar past.
But under G. Richard Wagoner Jr., the 49-year old former Duke University basketball player who became GM’s chief executive officer in June 2000, the company has begun a significant, technology-driven overhaul that is already paying dividends. As rival Ford fell from grace with a $5.5 billion loss in 2001, GM wowed Wall Street with surging earnings and several quarters of sustained growth. Excluding special charges, GM earned $1.5 billion in 2001 and nearly $2.3 billion in the first six months of 2002 alone.
Design and Conquer
While GM’s new cylindrical, steel and glass Renaissance Center tower in downtown Detroit may serve as headquarters for this revival, much of the company’s new ideas take shape at the technology campus in Warren. Down the hall from the Advanced Design Studio is GM’s Virtual Reality Lab, another dimly lit theater equipped with a wraparound, floor-to-ceiling display wall and special 3-D glasses that GM executives don at design reviews. This is several generations removed from the quirky 1950s version of 3-D projection. Here, a visitor receives a virtual look at the new Chevy SSR from various distances and angles. The digital images are so realistic that sunlight and trees reflect off of the SSR’s exterior. Visitors feel as if they can reach out and touch the front grill of the new pickup. Properly positioned, one can occupy the virtual space of the driver’s seat and get a sense of the interior or use a sophisticated joystick to tool the prospective vehicle through simulations of downtown Las Vegas or the Bonneville Salt Flats in Utah.
The Virtual Reality Lab, built in 1999 and now linked digitally to GM studios overseas, allows Lutz and his charges to collaborate on product reviews with colleagues across the world. These virtual reviews helped GM designers find an existing platform-the GMC Envoy sport-utility vehicle-upon which to base the SSR pickup, thus saving months and millions of dollars in building clay models. According to Szygenda, the SSR moved from sketch to production in less than 18 months. The entire project was done virtually, up until the construction of a show car.
With its shorter production cycle, GM is much more responsive to changes in consumer tastes. Five years ago, it lagged well behind industry averages in product development. A car slated for the market in 2000 had to be initiated in 1996 or earlier, forcing design committees to gaze into crystal balls to determine what their customers would want in five years. More often than not, they guessed wrong.
The combination of GM’s daunting bureaucracy and clunky manual product-development processes also spawned dated, unappealing vehicle design. Executives would often see just one design option for a prospective car or truck. If they recommended a change, it could take months to build a new physical model and get sign-off from manufacturing that the updated version was buildable. Most original designs were redone 50 times before they reached the marketplace, adding years to the process and rendering the initial concepts unrecognizable.
And if creativity is the end, the new technology may be the means. “The idea is to give flexibility to the designer on the front end and build technology that organizes the thought processes that permit creativity,” Szygenda says. Not only can designers display dozens of options to senior management, but they can also respond to data gleaned directly from dealers and customers via the Internet. By monitoring buying preferences on GM Buy Power, for example, engineers develop a far better sense of what customers actually want. The company has even experimented with the online design of a vehicle aimed at a younger audience into extreme sports. Iterations of the design were posted online, and prospective customers proposed modifications. While the car was never actually manufactured, designers were able to tap into the minds of potential drivers. “This is a lot better than trying to figure out 48 months in advance what you think customers will want,” says Robert M. Ottolini, GM’s executive director of engineering process and math strategy.
Actually building and testing those designs before they appear in showrooms is another reason for GM’s considerable investment in computer-aided engineering and simulation. The digitization projects are intended to overturn the ponderous “build, test, break” model that it had employed for decades. The goal is eventually to replace all vehicle dynamics and crash testing with digital simulation. Ten years ago, for example, 80 Chevy Caprice test cars had to be built and plowed head on into walls in order to ensure their crash worthiness and to meet other federal regulations. The process was time consuming and-at a cost of between $500,000 and $1 million per test vehicle-expensive. Today, with digital simulation replacing physical testing, the number of test vehicles is down to 10 or fewer, saving at least $35 million per product, despite a steady increase in the number of regulated tests required by the federal government.
“We wanted our engineers to spend more time analyzing vehicle dynamics, crash tests, handling, noise vibration and things like that rather than building models,” says Kirk Gutmann, GM’s global product development information officer. In order to support all this simulation, the company increased its supercomputing capability, provided by Hewlett-Packard and IBM, by 1,200 percent.
The result is unprecedented excitement and enthusiasm within the product development group. “When engineers are looking at 12-month design cycles, it’s a lot more exciting,” Ottolini says. “From the time you start working on a design to the time your neighbor can buy the vehicle is not that long. People like that.”
The Road More Wired
All this is aimed at making GM a smaller, quicker, more agile organization. The payoff has been double-digit productivity gains in each of the past five years.
“The real reason GM is so resurgent is purely a story about costs, operational efficiency and productivity,” says Baba Shetty, an analyst in the automotive practice at Forrester Research in Cambridge, MA. “How technology can help deliver on the core processes critical to the business-that’s where the real story is. Because even incremental improvements will pay off dramatically in the marketplace.”
The most notable thing at GM may be its reluctance to blow its own horn. After a tough decade in the 1990s, GM is well aware that the automotive industry is subject more than most to the swift business cycles of the economy. The company still has the challenge of developing distinct brands and brand values in the marketplace so that it is not simply competing on price. Though its technology initiatives have dramatically affected the bottom line already-by increasing productivity and earning good will among customers, dealers, suppliers and employees-everyone agrees there are more ways to use information technology to improve business.
“This will be more of a journey for the next five years,” Gutmann says. “There will be fewer big changes but more year-over-year improvements. We’re moving closer to real knowledge management, to sharing knowledge across the enterprise. Our goal is to make GM feel like a small company.”
Geoffrey Hinton tells us why he’s now scared of the tech he helped build
“I have suddenly switched my views on whether these things are going to be more intelligent than us.”
Meet the people who use Notion to plan their whole lives
The workplace tool’s appeal extends far beyond organizing work projects. Many users find it’s just as useful for managing their free time.
Learning to code isn’t enough
Historically, learn-to-code efforts have provided opportunities for the few, but new efforts are aiming to be inclusive.
Deep learning pioneer Geoffrey Hinton has quit Google
Hinton will be speaking at EmTech Digital on Wednesday.
Get the latest updates from
MIT Technology Review
Discover special offers, top stories, upcoming events, and more.