(FORTUNE Magazine) – There's a cute minicar called Ka zipping around Europe that wears a Ford badge on its grille. Yet much of its development was done by a three-year-old company you probably haven't heard of: MSX International, a privately held outfit in Auburn Hills, Mich. What the heck is MSXI? It is the biggest of the so-called outside engineering firms that are doing an increasing amount of work for the automakers, even to the extent of assembling vehicles.

Engineering services, which account for over half of MSXI's annual revenues of more than $1 billion, are growing by 10% to 15% a year. Outside engineering, says MSXI's high-profile CEO, Thomas Stallkamp, has become a "permanent feature" of the way the Big Three and other auto companies worldwide are doing business. Stallkamp is in a position to know. One of the early advocates of supply-chain management, he masterminded an overhaul of the way Chrysler deals with its suppliers, and rose to become vice chairman. He left last year in the management shuffle following Daimler-Benz's takeover of the company.

Other firms moving into the field include such fabled European marques as Lotus and Porsche. To garner more of Detroit's business, both are beefing up their for-hire engineering branches in the U.S. Observes Ron Harbour, president of the Harbour & Associates consulting firm in Troy, Mich.: "The auto industry is using a bigger share of outside engineering services than anybody realizes. What gets all the attention is buying parts on the outside. But the pressure continues on the automakers to reduce the head count in their salaried ranks, and outsourcing engineering is definitely a growing thing."

The practice is spreading to other manufacturing industries. A look at why provides insights into how managers running product-development programs are arriving at make-or-buy decisions not just for components but for design and engineering services as well. At Century Aerospace in Albuquerque, a startup developing a six-seat business jet that it hopes will be the lowest-priced on the market, money is the reason. There isn't enough of it to pay for a complete engineering staff. So a hired gun, using a university's wind tunnel, is in charge of critical model testing and aerodynamic refinement.

Bausch & Lomb Surgical in St. Louis (a unit of Bausch & Lomb, based in Rochester, N.Y.) had a different reason for looking outside. It wanted to incorporate a new type of pump module into an eye-surgery device it was developing, but nobody in the company's engineering group was sure exactly what type of design might do the job. Knocking on the door of a consulting firm led to an answer so clever that the consulting firm was granted three patents.

Paying a visit to MSXI is not a matter of driving up to a building and knocking on the door. The company employs more than 17,000 people in 24 countries. In the Detroit area, MSXI occupies two dozen buildings. This reflects both its rapid growth through acquisitions and a security consciousness that pervades this business. Asking outside engineering companies what they're working on is a lot like saying "What's new?" to a pal employed in a classified military program; the answers are of necessity vague. Operating in many buildings enables MSXI to shield the programs of secrecy-obsessed clients from the eyes of competitors. These may be other clients who are all too eager to learn what's in the other guy's product-development pipeline.

MSXI was formed in 1996 as a partnership between Citicorp Venture Capital and what was then the technical-services group of MascoTech. (The rest of MascoTech, which is headquartered in Taylor, Mich., now concentrates on making auto components and industrial products.) Through acquisitions, MSXI has entered businesses as diverse as providing companies with infotech staffing, furnishing people to man customer-service call-in centers, and managing the technical-document archives of clients such as Ford and Fiat.

The Ka program is the one big outside-engineering fish that Ford has given MSXI permission to brag about. In the early 1990s, when the European minicar market heated up with such hits as the Fiat Uno and the Renault Twingo, Ford got caught without an entry. John Risk, now president of MSXI's engineering unit, was the Ford manager in Europe who had to figure out how to bring a minicar to market fast. The automaker is now producing them at the rate of 150,000 a year.

"I went outside to do the Ka in 1996 because of both time and cost," Risk recalls. "We needed the car quickly. It would have been very hard to hire enough people in so little time, and the company didn't want to move people in from other important programs. So we used MascoTech [now MSXI]. They agreed to a fixed price to do the initial engineering around an existing Ford powertrain, execute the final design, line up the suppliers, and take it all the way through to production at a plant in Valencia, Spain. Today there are 25 MSXI people in Valencia working on quality improvements, cost reductions, and the next model cycle."

Richard Parry-Jones, Ford's group vice president for design and quality, sees a solid future for outside engineers. Says he: "There will be a very strong and steady market for engineering services in supplying flexible support to the automakers when they need it. On the Ka program, MSXI provided all the skills we needed, and worked with the Ford managers to identify shortfalls and fix them." MSXI also worked with suppliers, Parry-Jones notes. "Increasingly we've seen the engineering-services providers stepping in to help major suppliers who have needed to increase their engineering skills as they take on responsibilities for larger portions of vehicles."

On a recent visit to several of MSXI's Detroit-area facilities, this writer saw engineers in low-lit rooms peering at computer-aided design (CAD) terminals as they ran simulations of contemplated manufacturing systems or finessed detailed part designs for clients. In other areas, model makers were shaping clay mockups of future vehicles, and workers were checking out tooling and assembly equipment in preparation for shipments to customers' plants. Elsewhere, designers in a studio equipped with the latest in computer-graphics terminals were conjuring up vehicle shapes and the packaging of their internal components in rich detail.

A peek inside one well-secured MSXI building is enough to make a car junkie as happy as a UFO buff would be in a secret hangar full of alien corpses. It contains many of DaimlerChrysler's crown jewels: show cars, prototypes, and other hand-built wonders, including the Dodge Viper racecar that won in its class at Le Mans in 1998. As caretaker of these priceless vehicles, MSXI keeps them running and transports them around.

Often, speed is what saves the day in an engineering program. When necessary, MSXI can set in motion 24-hour engineering tag teams around the world, an arrangement in which each team at the end of its workday hands off work to the next group, thousands of miles away. "We have some Audi and VW programs in which we start work somewhere in Germany, then transfer the work here to Detroit, and then hand it off to Australia, so it never stops," says Risk. "This is done mostly in the CAD phase of body-engineering projects, where time is more valuable than money, and getting things fixed fast saves in total project cost by avoiding missed schedules."

Stallkamp thinks outfits like MSXI are ideally suited to handle so-called niche-vehicle programs. Such vehicles can require as much engineering as high-volume ones but are built only in modest numbers. He points, as an example, to the battery-electric version of the Ford Ranger, which MSXI engineered and assembles. A current project is a study of the feasibility of building a derivative of one of DaimlerChrysler's most recent models that wasn't in the original product plan.

"If I were an automaker," Stallkamp says, "I would put my engineering resources into developing most of the new stuff, because nothing makes you well again like new products. And I would let outsiders handle the maintenance and refinement of the current programs. Some of our clients have told us that they will permanently outsource a percentage of their engineering workload. Ford will never take it all back inside again, and Daimler's in the same category. Outside engineering is a permanent change in the way business is done. There's no manufacturing business that needs to be vertically integrated anymore. It just costs too much."

Niche vehicles are also a theme at Lotus Engineering in Ann Arbor, Mich., the contract-engineering branch of the famed British maker of sports and racing cars. Now owned by Proton of Malaysia, the company has a long record of building hot-rod versions of production cars, stretching back to the Lotus Cortina it built in 1970 for Ford of Europe. For the 2001 model year, GM's German subsidiary will roll out the Opel Speedster two-seater, a sporty all-new niche car designed and built by Lotus that has already been previewed on the show circuit.

Given its experience with its own sports cars, Lotus is an old hand at short development times and small production runs. This agility can be put to work on behalf of clients that want to develop and produce exciting variants of their standard offerings in a hurry. "Niche cars are going to grow heavily," says vice president John Briggs. "All of the major makers are bringing out products like the Ford T-Bird, and Chrysler's done it time and again with the Viper and the Prowler." Such cars, says Briggs, are an antidote to boring mass-market designs: "So many cars are coming out as computer-generated blobs. People are fed up with this corporate-average shape, and they want something just a little different."

Though much of its work is temporarily crammed into old factory buildings, Lotus will soon be able to stretch out. Construction is nearing completion on its new headquarters building in Southfield, which is not coincidentally located midway between key facilities of the Big Three. Lotus got its American foothold in Ann Arbor early this year when it acquired an engine-testing firm and its staff of 56. It's a logical fit with the company's strong background in powertrain development.

Lotus claims that engines bearing its engineering imprint power 10% of the cars sold today in Europe, and clients' engines and transmissions can be seen wailing away on dynamometers under computer control in the Ann Arbor facility's 26 glass-windowed test cells. The powertrain development and testing business has a bright future, thanks to ever-tightening federal exhaust-emission and fuel-economy rules. "A typical new-engine program now takes 50% to 100% longer than a vehicle program does because there are more legislative and durability issues to be resolved," says Lotus vice president Tim Holland. "Fully developing a powertrain can require 25,000 hours in the test cell."

When the Southfield building opens, Lotus will start with a staff of about 20 devoted to computer-aided design of engine and suspension components and body structures. Building and testing of prototype vehicles, as well as the powertrain group, will remain in Ann Arbor, whose industrial zoning permits dynamometers.

Certifying that a vehicle meets EPA pollution regulations can bring unusual machinery to your doorstep. Lotus has a motorcycle dynamometer, and atop it could recently be seen a hand-built bike powered by a huge Chevy V-8 engine of the sort normally found in a Corvette. Called the American King V and built by a guy in Louisiana, the V-8 bike needs an EPA smog check before it can be exported to power-crazed customers in Canada. It's not a typical job for Lotus, which normally works for big companies, but a colorful one. The thing rumbles like a hopped-up 1970s Camaro.

Lotus has a raft of proprietary technologies. It also has a low-volume production line in England, which can be booked for the assembly of clients' niche vehicles. Ford's Parry-Jones, for one, appreciates the availability of this option. "Engineering-services suppliers can provide a high degree of support to a particular boutique or niche-model program, up to and including manufacturing in some cases," he says.

Not all the contract-engineering firms' business comes from manufacturing companies. Once in a while a trade organization in search of a new technology will ask for help. Just such a situation brought a nice piece of work to Porsche Engineering Services in Troy, Mich., the U.S. arm of the German sports-car maker's consulting unit. The client was the American Iron and Steel Institute (AISI).

Suppliers of steel, aluminum, and plastics are locked in a perennial three-way battle to sell to automakers. Several years ago Ford was considering switching what it calls the closure panels on the Taurus--parts such as the hood, doors, trunk lid, etc.--from steel to aluminum. Faced with the threat of losing a big piece of business, AISI's automotive applications committee decided that it was time to shop around for ideas beyond those of its own engineers for using steel efficiently in car bodies.

A search committee was sent out to interview contract-engineering houses with body-building expertise. "We really liked what we heard from Porsche Engineering," says Peter Peterson, U.S. Steel's director of automotive marketing, who coordinated the effort. "They said we don't think you should do this program, because you want to take weight out of the closure panels and that isn't where most of the weight is." Instead, Porsche proposed a program that would rethink the way a Taurus-like unibody structure is designed and assembled and look for weight-saving opportunities.

AISI signed a $2 million contract with Porsche for an initial study, which the engineering firm proposed to follow with a $20 million project involving the construction and testing of full-scale bodies of advanced design. The mounting price tag sent Peterson out looking for more money, which he managed to raise by including among the project's sponsors 35 sheet-steel producers in 18 countries around the world. The group called itself the ultralight steel auto body, or ULSAB, consortium.

The four-year ULSAB program, which finished its work in 1998, had a spectacular outcome. Porsche's engineers came up with a body-structure design that weighed as much as 36% less than a benchmark group of nine production midsized sedans while costing somewhat less to build. And the

ULSAB structure met government-mandated crash-test requirements while providing far more stiffness than the benchmark cars. The release of the project's final report was a great day for the steel people. And through the next model change, at least, the Taurus will have an all-steel body.

Porsche engineers attacked the challenge from several angles. "What they did, architecturally, was to view the side of the body as though they were building a bridge," says Peterson. "And rather than have the bridge be nothing more than a deep box girder at the bottom, they ran an arch all the way from the A-pillar at the side of the windshield back to where the rear shock absorbers are attached. It's like the arch on a typical bridge from which mass is hung. This is the first major architectural change since the industry switched from body-on-frame construction to unibody designs back in the 1970s."

The strong, lightweight arch Porsche specified was made by a technique called hydroforming. In this room-temperature process, a tube of steel called a blank is placed in a mold. Fluid is pumped into the tube under high pressure, forcing it to expand and conform to the mold's contours. Parts for GM pickup trucks and the Cadillac Sedan De Ville are made this way.

Porsche's ULSAB body makes extensive use of two other advanced production methods. Long seams are formed by laser welding, which makes a continuous, stronger connection between panels than spot welding. The third method uses parts called tailor-welded blanks, consisting of pieces of sheet steel of differing thickness and alloy content that are welded together. Run through a stamping press, they form the lightest and lowest-cost body structures that can bear local stress loads.

Flush with its ULSAB success, AISI launched two other vehicle-engineering programs. One contract, awarded to Lotus, was for developing ultralight steel suspension parts. That one was sparked by BMW's use of an all-aluminum suspension in its 5-Series car, Peterson says. The other, awarded to Porsche, is a program to develop a lightweight steel vehicle platform ready to accept the sort of high-efficiency powertrains envisioned by the Department of Energy's Partnership for a New Generation Vehicle program, or PNGV.

Like a rival European program running on the same timetable, PNGV aims to develop an 80-mpg four-passenger car. Steel was originally excluded from PNGV's specifications in favor of more exotic materials, but AISI's lobbyists persuaded the program's overseers, the National Academy of Sciences, to consider steel as well.

Steering clear of the politics and keeping the design work on track is August Hofbauer, president of Porsche Engineering's U.S. branch, which opened in 1991. He's apologetic about not being able to describe exactly what, aside from the ULSAB follow-on program, his staff of 110 designers and engineers is working on, except to note that much of it relates to body engineering, sometimes for big trucks.

Porsche also has a styling studio with a staff of 25 in Huntington Beach, Calif., which, like the many other studios in that neighborhood, derives spiritual fuel from Southern California's exuberant vehicle culture. Porsche sends bigger jobs, like a crash simulation on a Cray supercomputer or the construction and testing of structures such as ULSAB bodies, to the company's extensive engineering complex in Weissach, Germany, where prototype vehicles can be built. There's even a low-volume production line in Weissach, left over from the days when racing rules obliged Porsche to build a certain number of special cars in order to qualify a model for competition.

Outside engineering firms, if anything, play an even more important role at small aircraft companies than they do in the auto industry. Instead of merely supplementing the client's engineering staff, they often are the staff. At Century Aerospace, a tiny ten-person firm camped in a hangar in Albuquerque, almost all the engineers are outsiders. The firm's founder is Bill Northrup, who dreams of building a cheap business jet. In his previous endeavors Northrup built an electric co-generation plant and ran a company that salvaged rails, ties, and ballast from abandoned railroad tracks. He has been a pilot since his teens. The Century Jet, now in the detailed design phase, is a six-seat, twin-engine craft intended to sell for $2.7 million, which would undercut the lowest-priced jets presently on the market.

Northrup launched his program in the early '90s, working with consulting engineers on the plane's prospective architecture. "You can't really afford to build a staff when you're a startup like we are, or you'd run out of money immediately," he says. "But you've got to get the guys who've been there and done it, so you don't repeat other people's mistakes."

After some early ideas for ultralightweight construction were discarded as impractical, Century ended up with its current cadre of engineering advisors, led by Dale Ruhmel, a veteran of 25 aircraft-certification programs. For the key low-speed flight control development and wind-tunnel testing phase of the work, Ruhmel turned to Paul Robertson, president of Aeronautical Testing Service (ATS) in Arlington, Wash.

A tall man with a dry wit, Robertson had worked during his days as a Boeing engineer on projects including the flap system for the 767 airliner and a gigantic windmill that needed aerodynamic tweaking. Feeling unsuited to the megacompany work environment, Robertson bailed out of Boeing in 1984. He started ATS as a provider of wind-tunnel testing services and also developed small-aircraft applications for vortex generators, little add-on fins that can improve airflow.

Using a big computer-controlled machining center, Robertson's shop carves wing and fuselage sections for his precisely contoured wind-tunnel test models from chunks of solid aluminum. The metal cutting is guided by a detailed CAD file describing the shape of the aircraft being tested. The Century Jet model, which has a wingspan of about seven feet, has been through three sessions in the University of Washington's Aeronautical Laboratory wind tunnel in Seattle, where Robertson leases test time. Strain gauges on the strut supporting the model sense the 700 pounds and more of lifting force its wings can generate when air in the tunnel rushes by at 150 mph.

Robertson's work focuses on measuring the overall efficiency of the Century Jet's shape, and in particular the efficiency of the flap system he has designed. The model has movable flaps and other flight controls, so test data can be gathered on the way they behave throughout their ranges of motion. The job of the flaps is to temporarily increase lift, letting the airplane take off and land at slower, safer speeds. The wind-tunnel program has now been through two model designs with two different wings, and things are looking good enough that it's time to proceed with construction of a Century Jet prototype for flight testing. Robertson will build the tooling needed to fabricate the flaps during the prototype's construction. "Aerodynamically and structurally, this airplane should work without needing a wave of Tinker Bell's wand," he predicts.

"I've been very fortunate to get guys like Paul Robertson onboard the program," says Northrup. "He's helped us with changes to the tail, the wing, and the flaps. He suffers no fools. He knows how to come up with something that can be certified and will fly as advertised."

The medical-equipment industry, too, is going extramural for engineering help. In this highly competitive field, gaining a clear technological edge can have a huge effect on sales. That's because doctors want to be able to offer the latest, greatest treatment for what ails you. It is against this backdrop that the engineers at Bausch & Lomb Surgical strive to come up with better equipment for the delicate practice of ophthalmic surgery. Reader warning: You may be about to learn more about cataract operations than you really wanted to know. But hang in, because what they do is clever.

In an eye that's "cataracted," the flexible, clear lens at its front has become hardened and cloudy, impairing the patient's vision. Removing the cloudy lens and replacing it with a plastic one is the cure. Bausch & Lomb embarked in 1995 on the development of a new modular eye-surgery system called the Millennium CX that can be used to treat both cataracts and disorders or trauma occurring in the deeper regions of the eye. Surgeons choose the desired modular features when buying the system, depending on the nature of their practice.

Of all the fields where design finesse comes into play, equipment used by surgeons has got to top the list for sensitivity to human-factors engineering, also called ergonomics. If the equipment doesn't feel precise and right to its user, the operation may not go as well. Consequently, medical-device developers spend a lot of time interviewing doctors and giving them prototypes to play with.

During a cataract procedure, the surgeon inserts an ultrasonically vibrating needle into the lens of the anesthetized patient's eye. Moving the ultrasonic tool across the lens pulverizes it, creating room for the plastic replacement. As the surgeon works, he or she flushes the area with sterile saline solution to remove the debris. Sometimes the doctor will also want to vacuum away excess fluid to keep the work area clean. In the past, Bausch & Lomb offered a pump working on a so-called peristaltic principle (as your digestive tract does) to flush, or irrigate, the area. And it used a second vacuum-type pump to suck away fluid. The two systems had different response times to their controls and felt somewhat different to use.

Talks with surgeons in focus groups had convinced the design team that the market would love a tool that both pumped fluid and vacuumed it up, and that could easily switch from one function to the other. "We didn't know exactly how to do it, but we knew we needed a system with two modes," says electronic-product design director Thomas Moore.

Earlier in his career, when he worked in electric-power systems engineering, Moore had become familiar with Arthur D. Little, the consulting firm in Cambridge, Mass. So he called to see if ADL had any ideas about how to approach the push-pull pump question. A week later he got a response from Bob Farra, a director in ADL's technology and innovation business unit, who thought that a "scroll" pump might do the job. This unusual device consists of two spiral, or "involute," components that resemble a cross section of a nautilus shell. One involute remains stationary, while another interleaved with it traces a gently orbiting path of motion against its mate, squeezing whatever is in the pocket of space between and chasing it to an exit.

Among the characteristics of scroll pumps are efficiency, quietness, and reliability. They can be designed to run in both pumping and vacuuming modes, switching from one to the other when the drive motor is reversed. Scroll pumps are used in large numbers as compressors for air conditioners in homes and vehicles, and have also been used to supercharge car engines. ADL already had close to 30 patents related to the devices, and derives about $1 million a year from licensing them.

In addition to combining pumping and vacuum functions, Bausch & Lomb required that the system be disposable, to eliminate the cost and contamination risk of sterilizing components that come in contact with patients' body fluids. Farra's group agreed to construct a prototype system for evaluation. Its scroll-pump components were machined from stainless steel, which would be far too costly for a production item but was perfect for testing. It was tried out with a hollow ultrasonic cataract needle prepared by Bausch & Lomb, which can deliver a saline solution by letting it flow down its outside or suck away fluid and debris through its interior. Surgeons who tried the device loved the responsiveness and feel of the two-way pump feature, which is operated by a foot control.

ADL agreed to proceed with development of the system, but not just as a chunk of contract engineering. Rather, the consulting firm licensed the technology exclusively to Bausch & Lomb for use in eye surgery. Bausch & Lomb considers it a win-win deal. ADL is free to license the pump only to other industries and only for other uses, and has a vested interest in the success of Bausch & Lomb's pumps. Notes Moore: "They get royalties on every cartridge sold."

ADL's six-member development team consisted of specialists in electrical and mechanical engineering, software, human factors, and manufacturing. As it turned out, the firm was awarded three patents for inventions during the program. One was for a cheap and reliable sensor that measures vacuum levels so that the system's computer can adjust the pump motor's speed as needed. The sensor consists of a diaphragm of thin stainless steel that flexes slightly in response to vacuum. The diaphragm's motion relative to a stationary printed circuit is read as a variable capacitance, which the controller translates into pump-speed commands.

Making the dual-mode system cheap enough to be disposable involved the development of a scroll pump made from injection-molded plastic, the first of its type. Here the challenge was in the details, such as designing in a slight amount of interference, or rubbing, between the involutes to ensure a tight seal. Another challenge was locating a specialized supplier that could meet the dimensional tolerances and cleanliness standards required for a medical device.

The end result is a disposable plastic Concentrix cartridge that sells for less than $100. The surgeon just pops it into a slot in Bausch & Lomb's Millennium system when preparing for a cataract operation. An external motor drives the scroll pump inside the cartridge through a connection that's isolated from the fluids within. When a used cartridge is tossed out after each operation, so are any worries about contamination.

Moore thinks that farming out this specialized piece of engineering was a smart move. "We pride ourselves on our core competencies like developing software and ergonomic ultrasonic hand pieces," he says. "We've gota team of about 50 people in our design and development group doing these sorts of things, which we want to keep in-house." But, in words that echo those of other companies turning to contract engineers, he adds, "If we need something else, we will go outside to find it."

FEEDBACK: sbrown@fortunemail.com