Skip navigation

RETHINKING Platform Engineering

The idea looks great on paper and wows 'em on Wall Street: spin off dozens of different vehicles using the same basic underpinnings and 60% or 70% common parts. The vehicles look very different on the outside, but underneath and under the hood they're very similar. Oh, and by the way, you can build each at just about any assembly plant, anywhere in the world. The concept often referred to as platform

The idea looks great on paper and wows 'em on Wall Street: spin off dozens of different vehicles using the same basic underpinnings and 60% or 70% common parts. The vehicles look very different on the outside, but underneath — and under the hood — they're very similar. Oh, and by the way, you can build each at just about any assembly plant, anywhere in the world.

The concept — often referred to as platform engineering — can slash vehicle development times and radically chop design, engineering and manufacturing costs. It has become the auto industry's Holy Grail as it seeks to reduce expenses, improve quality and deliver a wider variety of products to increasingly discriminating consumers in a variety of world markets. And it has spawned a new generation of highly flexible assembly plants that can pound out lots of different cars and trucks, instead of being dedicated to just one.

Only a few are mastering all the elements, but all are going after the basics. Just about every major manufacturer has vowed to slash the number of car and truck platforms it uses over the next few years. Among them:

    General Motors Corp., the world's largest automaker, is reducing the number of its global car platforms or “architectures” as it prefers to call them, from 13 to 7 by 2005. (It will keep all 6 of its light truck architectures through 2005.)

      Ford Motor Co., which has an estimated nine car and 10 truck platforms (including Volvo and Land Rover products) also is consolidating, although it won't say by how much. Ford products such as the Lincoln LS and its European Mondeo currently share platforms with Jaguar, and the next generation Ford Taurus reportedly is based on a Volvo platform.

        Industry sources say Toyota Motor Corp. is in the process of reducing the number of platforms it uses from about 40 to about 14, even though it is considered one of the world's best at platform sharing (see Corolla story, p. 53).

          At DaimlerChrysler AG, Mercedes and Chrysler products won't share platforms, but Chrysler probably will share with DC affiliates Mitsubishi Motors Corp. and Hyundai Motor Co. Ltd. Michael Robinet, director of forecast services for CSM Worldwide, says Chrysler's LX platform, due out in '04 or '05, may be the last unibody car platform developed by the Chrysler Group. Later car platforms likely will come from Mitsubishi or Hyundai. Industry sources say DC's Korean partner already plans to slim down from an estimated 23 platforms to six or seven.

            Renault SA and Nissan Motor Co. Ltd. are whittling their platform portfolio down from 39 to 10 (see sidebar, p.49).

              Volkswagen AG, considered by most experts to be the world leader in platform consolidation (although it isn't nearly as efficient at manufacturing them as Toyota and Honda Motor Co. Ltd.), already has chopped the number of its platforms from 16 eight years ago to just four now.

                “We're far from the point where we've seen the total optimization of platforms on a global basis; a lot of consolidation has only started occurring in the last three or four years,” says Jim Mateyka, vice president of the Global Automotive Practice of A.T. Kearney. “We're just beginning the process.” With only a handful of automakers controlling 80% of global automotive production, the trend toward fewer, more flexible, higher-volume platforms is inevitable, he says.

                CSM's Mr. Robinet predicts average production volume per platform likely will rise 30% by 2010 and many purely North American platforms will disappear in favor of global ones by 2010.

                However, John E. Cunningham, a partner with the global automotive practice of Accenture, formerly Andersen Consulting, points out that there always will be some strictly regional platforms, because it doesn't make sense for some — such as fullsize pickups — to be designed for global markets.

                But even as the industry plunges headlong into platform consolidation, a dark side to this trend is emerging. It has caused Volkswagen to stop in its tracks and switch direction while others watch nervously from the sidelines. The hard truth is that VW's much ballyhooed platform strategy hasn't met financial or sales expectations.

                What's more, savvy European car buyers view VW's platform strategy as a shell game. Instead of paying top dollar for the German automaker's best models, they're trading down to cheaper — but technologically similar models — based on the same platforms but manufactured by VW's lower-price brands, Skoda and Seat.

                VW's products currently share about 70% of their parts, causing some to speculate that engineers tried too hard to commonize. “That may be pushing the envelope,” acknowledges David E. Cole, director of automotive research at the Environmental Research Institute of Michigan.

                Just what exactly is a platform? Nailing down an apples-to-apples definition between manufacturers isn't easy.

                Before it decided to change strategies, Volkswagen defined a platform as being comprised of the front and rear axles, steering components and engine, side members, floor pan, and the fuel tank. Each body built on the platform is marque-specific and referred to as a “hat.” Production of platform and “hat” components is commonized and sourced globally. Only 10% of the 4 million vehicles produced by VW in 1996 were based on the platform strategy, now, almost 100% are.

                VW now is moving away from a platform-centric approach toward a component-sharing strategy, which some insiders believe will take the focus off “platform hard points” and allow more differentiation between the brands.

                VW reportedly is focusing on 11 key modules, including engines and transmissions, brakes, axles and fuel systems, with a target of completely transitioning away from platforms by 2005.

                Whether this constitutes a true “abandonment” of its platform strategy or just rethinking the concept and developing a new, more flexible type of platform remains unclear — and unexplained. VW, like GM and Ford, didn't respond to interview requests for this story.

                However, it's understandable that GM would just as soon forget the word “platform.” It was devastated by a platform strategy in the mid-1980s and early 1990s that led to all of its midsize cars being labeled as drab cookie-cutter “lookalikes” and it has spent years living that image down. Instead GM officials like to use the word “architecture” and relegate “platform” to a smaller role.

                According to GM, architecture is: “a set of common product and manufacturing standards related to a vehicle type (e.g. full-sized trucks).”

                At GM, a “platform” is “the application of those architectural standards to a related family of vehicles (e.g. full-sized SUVs).”

                Platforms also once were thought to be defined — and limited — by their width, which corresponded to the width of the assembly line they were built on. But Honda changed that idea in 1997 when it introduced the completely overhauled '98 Accord sedan on a “flexible” platform. The unusually flexible configuration gave birth to three distinct Accords that cost 20% less to bring to market than the single Accord developed four years before.

                It allowed Honda to introduce a big, truly midsize sedan for the North American market, plus two much narrower, shorter versions tailored to the needs of Japanese and European buyers. In essence, it created the answer to developing a truly global car. Rather than trying to develop a cookie-cutter car that tries to be all things to all markets — which rarely succeeds — it has a cost-efficient foundation for supplying a variety of different products customized to the region.

                That's the type of flexibility that everyone is trying to emulate, at least to some degree. But CSM's Mr. Robinet points out that while a lot of companies can build a variety of products off the same platform, only the Japanese seem to be able to master building a number of different platforms on the same line.

                Want further evidence? Honda is the first in North America to build an SUV and minivan on the same line (the Acura MDX and Honda Odyssey).

                However, Toyota isn't a slouch at this business. If you want to see how a platform strategy is supposed to work, all you need to do is take separate spins in Toyota's Highlander crossover vehicle, Sienna minivan and Avalon sedan. Or stroll over to your local Lexus dealer and drive an RX300 crossover vehicle or its sibling for the Japanese market, the Harrier. They're all heavily based on Toyota's bread-and-butter Camry sedan, as well as numerous other Toyota products offered in markets around the world. While each model is distinctive, they all share very similar lines and dimensions.

                Their basic architectures are built around the same specific points in space — known as hardpoints. That allows them to share expensive key components such as engines, instrument panels and suspension parts — and it usually enables them to be built on the same assembly line as their siblings.

                Camry, Sienna, and Avalon all are built on the same line at Toyota's manufacturing complex in Georgetown, KY. Highlander, RX300 and other Camry variants could likely be built there as well.

                If you want to see how platform engineering is not supposed to work, ask manufacturing engineers at DaimlerChrysler's woefully underutilized Neon subcompact plant in Belvedere, IL, why they can't build the hot-selling PT Cruiser, even though it's based on the Neon platform. The PT Cruiser reportedly is too tall for Belvedere's aging assembly line. Its “hard points” weren't designed with Belvedere in mind.

                But even while they slash the number of their platforms, automakers talk of sharing even more, between brands and between foreign partners (see sidebar).

                Now just when you would expect VW to start sitting back and reaping the rewards of its brilliant platform strategy, VW Chairman Ferdinand Piech has announced the company is shifting gears again. Instead of bragging about VW's leadership, he's eschewing the term “platform” for “modules” and promises this new strategy will yield bigger, better benefits than its previous course.

                Critics say VW may have outsmarted itself. Its cost-cutting strategy to trim the number of platforms down from 16 eight years ago to just four today has backfired on the German automaker in much the same way that it did for GM in the 1980s with its look-alike cars that crossed divisional lines. VW is suffering a buyer backlash of its own for spinning VWs, Audis, SEATs and Skodas off identical architecture. Its small A-platform, for example, is used on no fewer than eight models, from the VW Beetle and Golf to the Audi A4 and TT, Skoda Octavia and SEAT Toledo.

                The game plan is being blamed for a softening of VW sales, particularly in the home market of Germany, while deliveries of less-expensive but technologically identical and stylistically similar Czech-built Skodas have boomed.

                Volkswagen-brand car sales rose a scant 2.3% last year to 3.13 million units, and Audi sales climbed 3% to 653,404. At the same time SEAT deliveries jumped a more resounding 6.9% to 515,000 and Skoda sales shot up 13% to 435,403 — including a 19.2% rise in markets outside the Czech Republic.

                To make matters worse, the platform-sharing strategy hasn't translated itself to the bottom line as much as VW may have anticipated. In the late-1990s, VW Chairman Ferdinand Piech had promised the plan would save some DM3 billion ($1.4 billion) annually in engineering and production costs.

                But today, the automaker's profit margin is an estimated 3.5%; trailing its European rivals PSA Peugeot-Citroen's estimated 4.6% and Renault SA's projected 6.3%. That puts VW's profit-per-car at around $373, compared with Renault's $571 and Peugeot's $509, based on published estimates by Schroder Salomon Smith Barney.

                Mr. Piech has been quoted as saying there will be some 67 new VW Group vehicles launched within the next five years, including the new D1 that will bow later this year to take on BMW and Mercedes, and the new small Bentley and high-powered Bugatti that will compete in upper-stratosphere price niches.

                Also said to be on the drawing boards are five minivans — three from VW and one each from SEAT and Skoda. The product program assault reportedly will cost the automaker E20 billion ($18.1 billion) to develop and another E11 billion ($10 billion) in plant tooling.

                The new design philosophy presumably will allow different brand teams to choose from among the various component sets, paving the way for a SEAT Toledo to offer different powertrain combinations from the VW Bora (Jetta) and Skoda Octavia, for example.

                The first concrete example may be the next-generation Passat, which reportedly will break away a bit from its Audi A4 twin by going to a transversely mounted, rather than longitudinal, engine and sharing more modules from the Golf/Jetta line. For instance, Mr. Piech has said using the Golf brakes and power steering will cut E100 ($91) off the cost of the Passat, and switching to the Golf's transaxle will cut another E125 ($113) in cost.

                The strategy reportedly is seen saving VW up to DM6 billion ($2.8 billion) per year. Mr. Piech puts the savings at a more conservative E500 ($450) per vehicle. And he claims it will cut design time by 30%.

                Part of that efficiency gain could come from improved manufacturing flexibility under the new game plan. Productivity hasn't been a VW hallmark, particularly in Germany, where the automaker's Wolfsburg plant cranks out only 40 vehicles per worker per year, a far cry from the industry-leading 112 per employee at Nissan Motor Co. Ltd.'s operation in Sunderland, UK.

                Of course, skeptics contend that the switch away from “platforms” to “component sets” may be as much about semantics as engineering. Investors love to hear talk of platform sharing — maximizing the use of engineering resources and sharing parts across lines cuts costs, and that's music to the ears of market analysts.

                But consumers are becoming savvier, and some are asking, “Why pay more for an Audi when it's the same underneath as the far less expensive Skoda?” So putting the word out that there's a shift away from common platforms to more strategic component sharing may be the industry's way of trying to have its cake and eat it too.

                But some analysts play down the size of the negative impact in the market from sharing platforms across brands.

                “There is some (cannibalization of the VW brand), yes, but it is in the single digits,” says Karl Ludvigsen, chairman of Ludvigsen Associates U.K. Ltd., a publishing and consulting firm. “Most people that are buying a Skoda are buying it because VW is involved, and it's a better car than before. People don't compare that many cars; they don't look at all the alternatives. People don't like to experiment that much.”

                He says other brands — such as Peugeot and Citroen — also stumbled a bit when first going to common platforms. “But they learned from it,” he says. “PSA now is good at using common components and still making Citroen and Peugeot cars look different.”

                He says Ford and Jaguar also have done an excellent job, with Jag “shrewdly” launching its new Mondeo-based X-Type as a 4-wheel-drive only. “They'll offer a cheapo front-wheel-drive model later, but launching it as a 4wd first will help put distance between the Jaguar and the Mondeo,” he says.

                The real test in whether this new philosophy will yield widely different designs may be in whether stylists will be allowed flexibility with the one component module that commands constant scrutiny by all drivers: the instrument panel. “The IP still strikes me as the place where most companies don't want to invest any more than they have to,” says Mr. Ludvigsen.

                But Pandora Ellison, director of advanced programs at Delphi Automotive, a major instrument panel supplier, says IPs can have standardized structures and dimensions and still have very different “faces” that establish specific brand identities with different designs, gage packages and materials.

                The new plan that VW envisions already is in place at BMW AG, which is getting ready to spin some more magic off its highly utilized 3 series building blocks. Plans include a new sports/activity vehicle below the X5 in size and price, plus a sub-3 series car.

                The new sub-3 will make extensive use of existing 3 series components, possibly including powertrains and electronics.

                “Each of our products has the same profitability hurdle rate,” explains Helmut Panke, BMW board member in charge of finance. “And we can achieve that by making component sharing a part of our strategy. For example, we use our 6-cyl. engine in our 3-, 5- and in Europe our 7-series.

                “Thus, we suddenly have a very expensive component, including the drivetrain and gearboxes, across three completely different model lines,” he says. “The same can apply for the navigation system or onboard electronics. Thus we get into volumes for those very expensive components that make it possible to build a BMW-positioned product at a price point in the market that is acceptable and still makes a profit.”

                Hide comments

                Comments

                • Allowed HTML tags: <em> <strong> <blockquote> <br> <p>

                Plain text

                • No HTML tags allowed.
                • Web page addresses and e-mail addresses turn into links automatically.
                • Lines and paragraphs break automatically.
                Publish