The Snap-Together Business Jet Bombardier's new recipe: a dozen big pieces, four days to assemble them, and it's ready to fly.
By Philip Siekman

(FORTUNE Magazine) – The flags hang high above the assembly line, starting at the rear with the Stars and Stripes and progressing forward through Canada's Maple Leaf and the banners of Germany, Austria, Australia, France, Britain, Taiwan, and Japan. Down below, a new corporate jet is taking shape as it moves through five production stations in which great chunks of it are rolled in and joined together. The flags signal the sources of those big subassemblies: the engines from Phoenix, the nose and cockpit from Montreal, the mid-fuselage from Belfast, the tail from Taichung, the wings from Nagoya, and lesser parts from cities in four other countries.

For aviation buffs, this factory on the edge of Mid-Continent Airport in Wichita is hallowed ground. For nearly four decades, Learjets, the sports cars of corporate aviation, have been built here from sheets and bars of aluminum. Three Lear models still roll out of the facility. But the plane taking shape beneath the flags is the Bombardier Continental, named for Lear's Montreal owner and the jet's Miami-to-Seattle range. It's put together the way Bombardier will build all its planes in the future. Not counting rivets, it takes just a dozen big parts, all made elsewhere, to assemble the Continental to the point that it looks ready to fly--fewer than are found in a Revell model kit.

Although the first Continental won't be delivered until next winter, Bombardier has orders for 95 and has earmarked another 25 for FlexJet, a subsidiary that sells fractional shares of business aircraft. Since the Sept. 11 terrorist attacks, executive-jet sales have weakened, but no Continental customer has canceled. Three of the planes are now in flight test, and others are moving slowly through assembly. When the line is up to full speed, production time will rival that of an adolescent model builder: just four days to put together a plane and get it in the air.

Building planes with major subassemblies is not new. Boeing's Wichita plant fashions 737 fuselages for an assembly line in Renton, Wash. Airbus, which began as a collection of independent companies, makes major subassemblies in four countries and assembles in two. Executive-jet companies, however, generally make major parts in-house. Cessna, the volume leader, remains almost totally integrated vertically.

No aircraft company is as committed as Bombardier (pronounced BOM-BAR-dee-AY) to dealing out orders for major parts of its planes to distant factories inside and outside the company. No planemaker relies as heavily on outside suppliers for design support and the sharing of development costs and market risk. Bombardier hasn't arrived at this point without pain. Construction of a new Learjet, the Model 45, bogged down when it was parceled out internally among three company locations. Undeterred, the company fixed the problems and, on its next new executive jet, the long-range Global Express, expanded the approach by enlisting outsider Mitsubishi to build the wings.

The Continental is the most ambitious project so far. Bombardier is using two of its factories plus Wichita and a company-owned completion center in Tucson that will paint the planes and outfit the interiors. It has about 30 prime suppliers, including a dozen that have been involved since the initial design phase. While Bombardier has invested some $250 million in the Continental so far, its partners have probably matched that dollar for dollar. The outsiders also share the risks. Says John Holding, 58, who is in charge of Bombardier's engineering and product development: "They haven't got a contract that says, 'You're going to sell us 25 wings a year for the next ten years.' If the market's there, it's there, and if it's not, it's not."

As that extensive supplier involvement indicates, Bombardier is, in corporate-speak, more integrator than assembler. It doesn't simply tack together stuff made by others to its specifications. It manages a multinational, multicompany endeavor. As a corporation, Bombardier has also moved from assembly to integration, tying together a three-country collection of independent, and once wobbling, aircraft builders into what is now the third-largest aerospace company after Boeing and Airbus.

That's a long way from l'auto-neige, the snowmobile, with which Joseph-Armand Bombardier started the company in the '40s. In time his son-in-law, Laurent Beaudoin, now 63 and still chairman, took over, licensed French know-how, and in 1974 started making rolling stock for the Montreal subway. In 1986, Beaudoin bought another Montreal company, the government-owned planemaker Canadair, and went on to buy other aircraft builders with romantic histories and gothic balance sheets: Northern Ireland's Short Brothers in 1989, Lear in 1990, and, in 1992, Toronto-based de Havilland, then owned by Boeing.

The $10.8-billion-a-year company still makes snowmobiles, plus jet skis and other recreational contributors to noise pollution. Since its acquisition of DaimlerChrysler's rail equipment business last May, the company has become the world's largest manufacturer of things that run on steel wheels. But aerospace is what counts. Bombardier now claims about half the market for regional airliners and nearly 30% of business-aircraft sales.

In the fiscal year ending Jan. 31, 2001, Bombardier shipped 129 Lears, 74 other executive jets, 52 turboprop airliners, 105 regional jets, and ten amphibious turboprops sold mainly as firefighting water bombers. That brought in 66% of the company's total revenues, and more than 85% of its $644 million in after-tax profits. After rising 20% in the first nine months of the current fiscal year, the aerospace unit's sales have slowed. The regional-jet business sagged in September and has mostly recovered, but buyers of business jets, particularly smaller models like the Lears, have been holding off. In December, however, Bombardier got a big boost when Avolar, United Airlines' entry into the fractional-business-jet sales market, placed a $632 million order for 57 Learjets.

It's an aerospace axiom that executive-jet customers trade up through models progressively bigger, faster, and able to fly farther. So although it once seemed that Beaudoin couldn't pass an aircraft factory without buying it, the Lear deal was justified as a way to offer starter jets to customers who would later buy a bigger Bombardier plane. But Lear badly lacked cash and confidence, having gone through a succession of neglectful owners. In 1989, the year before Bombardier took over, it shipped just 13 planes.

Under Bombardier, Lear polished its aging line, bringing out two upgrades, a small Model 31A, and a mid-size Model 60. It then started work on a plane between them in size, the Model 45. The first "clean sheet" Learjet since the plane was introduced in 1963, the 45 was also the first effort by Bombardier to integrate its production capabilities by building the wing at de Havilland in Toronto and the fuselage at Short in Belfast, with final production in Wichita.

The Model 45 has been a sales success; Bombardier has shipped more than 180. Production was another matter. Inevitably, flight tests of any new plane reveal something that ought to be fixed on completed or partly built planes destined for customers. When that happened with the Model 45, recalls Tom Hilpert, the program director for the Continental, "the Learjet engineering team solved the problem and fed their solution back to Short or de Havilland. When the change got to them, they couldn't work with it because it called for different specs, different materials, different fasteners." Wings that arrived "stuffed"--that is, ready to mount--had to be unstuffed and rebuilt. At one point, 27 Model 45s were scattered about the Lear factory in various stages of completion while customers waited.

The headaches in Wichita extended to the rest of the plant, where the 31A and 60 are still built from aluminum stock up. In 1968, 1,067 Lear employees turned out 46 planes. Thirty years later 3,950 people shipped 64. That's when Bombardier parachuted in help: a new general manager, Jim Ziegler, now 44, who had served as a beleaguered treasurer of Lear prior to the acquisition and was then running Bombardier's customer service operations. Ziegler raided the rest of Bombardier for new management and set out to change procedures and attitudes. With little change in head count and lots of changes in the plant, production went up to 109 planes in 1999 and 129 in 2000.

The Lear 45 problems, recalls Holding, "were one of the things that drove us to recognize that we had to bring Bombardier Aerospace together as one group." It wasn't easy. Each entity took pride in its ability to take a plane from idea to air. Lear had its jets. De Havilland concentrated on turboprop airliners. Canadair worked on a wide-body executive jet, the Challenger, which it still sells and from which Bombardier's line of regional jets was derived. But gradually the aerospace group has become a single company. One sign of change: Increasingly it's "Toronto" instead of "de Havilland," or "Belfast," not "Short." "Lear" is still a brand name, but otherwise it's "Wichita."

Though the Lear 45's production difficulties were finally resolved within the company, Bombardier had no choice but to extend the subassembly approach by bringing in outsiders. When design engineers proposed spending nearly $1 billion to develop what would become the Global Express, Beaudoin told them to come up with a "better business proposition." When they returned, they had a deal with Mitsubishi, which had agreed to cover its own development costs and build the plane's wings as well as the fuselage segment to which they are joined. With that crucial part covered, the other major jobs were parceled out internally on the understanding that, this time around, if there was a problem with a part, those who built it, fixed it.

Bombardier has also learned to look to suppliers for solutions, not just hardware. Landing gear is bought as a system rather than as separate pieces. Engine builders provide a complete package from turbine to thrust reverser to nacelles. The aim is to bring it in and bolt it on. It's not that simple yet, but it's getting there.

While production of the Global Express was getting under way, Bombardier started talking to potential partners about the Continental, a completely new, "super midsize" business jet that could comfortably fly eight passengers nonstop from coast to coast. That puts it in a crowded niche that includes Cessna's $18.6 mil-lion Citation X, the French-made three-engine Dassault Falconjet 50EX, with an $18.8 million price tag, the $20.1 million Gulfstream 200, made in Israel, and Raytheon's $16.2 million Horizon.

In the past, says Victor Valenti, a marketing director at Honeywell Aerospace, sellers of business planes have tried to attract customers with "lots of emphasis on technology, on best performance, on the latest advances." However, the idea behind the Continental was to win the hearts of CFOs. Honeywell, chosen to build the engine, went for simplicity, reliability, and more thrust for the dollar. For the plane itself, says Mark Doyle, the parts logistics manager, Bombardier took "proven design technique and improved upon the manufacturing process to minimize the cost." Customers for the first 30 got an early-bird special: $14,250,000. Today's price is $15.1 million, still well below the competition's.

The Continental team now includes Bombardier Montreal for the cockpit and forward fuselage, Belfast for the center section, Mitsubishi for the wing, and a newcomer, Taiwan's Aerospace Industrial Development Corp., for the rear, including vertical and horizontal stabilizers. Initially each supplier dispatched teams of engineers to Montreal; at the high point there were about 250 outsiders and 250 from Bombardier. Says Bassam Sabbagh, now director of operations on Wichita's Continental line: "We looked at every part of the aircraft and got agreement on whether it was going to be manufacturable and whether the design was as good for, say, Mitsubishi as for us as the integrator."

The partners also put a lot of initial effort into trying to avoid surprises during flight tests. One concern in new aircraft is whether pieces of electronic equipment, including not only the avionics but also lights and sensors that are found just about everywhere on a plane, will interfere with each other after installation. To avoid that, Rockwell Collins, chosen to supply most of the avionics gear, built a system-integration test rig at its Cedar Rapids, Iowa, headquarters. It connected all the electronics and included real hardware such as the throttles. For eight months the other suppliers made the trip to Iowa to make sure their stuff worked with everybody else's.

While the designers labored, Jim Ziegler and his managers campaigned to get the assembly work for Wichita. They pointed to the turnaround in production and proposed to put the Continental together in existing space, partly by relocating a machine shop. In mid-1999, Wichita got the job. On Sept. 1, 2000, it began putting together the first plane.

Compared with a Lear, the Continental is big. Nearly 69 feet long, with a wingspan of almost 64 feet, it's more than 40% longer and wider than a Learjet 31A. Despite that, the assembly area is just 36,000 square feet. There's only one piece of bolted-down tooling: a rig where the mid-fuselage is mounted while laser-guided jigs move the forward and aft sections into place, hold them steady as rivet holes are drilled, back up so burrs can be cleaned off, and then return the huge parts to the identical spot for final joining.

The other crucial fit is when the fuselage is raised and the single-piece wing is slid underneath. They're joined by just three stainless-steel, high-strength bolts, which together can be held in one hand and which slip in without a jiggle. The two biggest are seven-eighths of an inch in diameter and 3.7 inches long.

After partnership programs on four Bombardier planes, including two regional airliners, the people fit together as smoothly as the components. A measure of that is the decreasing time between program launch and first delivery. The Lear 45 took 71 months; the Global Express 67. The Continental is expected to take 41 months. Sabbagh says the technology, the design, and the interfaces all get better from program to program.

The system still needs tweaking, in Sabbagh's view. Some parts or pieces that could be part of the subassemblies arrive at the line separately because they can't be installed until later. A part might be left off the fuselage, for example, because it would get in the way when the wing is being installed. Sabbagh says, "It all comes down to initial planning. I'd spend more time on the interface between partners and design out that sort of thing."

There's no doubt that Bombardier will continue the partnership approach. When a plane's development cost approaches $1 billion, it can't afford otherwise. But the question for any company outsourcing chunks of a highly technical product is at what point to draw the line. When is control lost because the company's knowledge of the things that distinguish one product from its rivals has become outdated?

John Harding says there's a need for in-house expertise: "You can only manage a partner if you know what you're talking about. You have to have people who understand hydraulic systems and avionic systems--not to the detail of knowing exactly how you machine a valve, but you have to have the knowledge."

Bombardier still keeps a hand in basic production. It builds wings for its regional jets and has a core competency in cockpits. Yet a lot of know-how is shifting to the partners. The best answer may be to make the relationship so seamless that each is, in some way, dependent on the others. Harding is now talking to a few partners about "the next phase, an even more integrated approach involving people earlier in the process." What he seems to be reaching for is a partnership of equals developing a plane for which almost every part from nose to tail cone is specifically designed to work with and complement every other part. He doesn't know exactly how that can be accomplished. But he is certain of two things: "I know we can do better. And we will."

They'll have to. Bombardier Aerospace just got a new boss, 39-year-old Pierre Beaudoin, the chairman's son. He plans to double revenues in five years while lifting pretax earnings at a compound rate of 20%.

FEEDBACK: psiekman@intrex.net. Stories from the Industrial Management and Technology section can be found at fortune.com/imt. Executives in manufacturing and research and others eligible to receive FORTUNE's Industrial Edition can subscribe by calling 888-394-5472.