Nearly seven years ago, when I visited Boeing’s cavernous manufacturing site in Everett, WA, the sight of machinists playing ping-pong in a vast but idle shop seemed to symbolize the stagnant state of the aviation industry. Air travel had not recovered from the terrorist attacks of September 11, 2001. And Boeing was facing stiff competition: Airbus, its European rival, had made innovative advances in commercial jets, such as rear tail pieces made from lightweight composites. Worse, Airbus was gearing up to build the A380 superjumbo jet–a higher-capacity, more efficient competitor to Boeing’s iconic 747.
Boeing needed to do something bold. So it bet its business on a medium-sized advanced aircraft called the 7E7–today known as the 787 Dreamliner–that would be 20 percent more fuel-efficient than other jets of comparable size and cost less to maintain. Such a jet would make direct flights between far-flung smaller cities (say, Boston and Bangalore) cost-effective. “It’s the future. It really is,” Mark Jenks, a Boeing vice president who was then director of technology integration for the 7E7 program, said to me in 2003. “If we get it wrong, it’s the end. And everyone here knows that.”
Boeing’s plane would greatly increase the use of advanced composites–layers of carbon fibers embedded in epoxy resin to form durable, lightweight materials. The 787 structure would be 50 percent composite, compared with just 12 percent in Boeing’s previous jet, the 777, and 23 percent in the Airbus A380. For the first time in commercial aviation, the entire tube of the fuselage would be a single piece of composite, replacing the customary aluminum alloy skin affixed to aluminum alloy ribs. In another first, the wings and the center wing box–a chunk of fuselage to which the wing structures attach–would also be made of composites.
Innovation extended to the design and manufacturing process. Airbus and Boeing had long subcontracted some manufacturing, and Airbus had even invited some subcontractors to invest and share the risk in the A380. But in an effort to reduce its own investment and cut costs, “Boeing took it a lot further,” says Hans Weber, owner of San Diego-based TECOP International, a technical consultant to the aviation industry and government agencies including the Federal Aviation Administration. Airbus had never outsourced design or the manufacture of the main airframe. With the 787, Boeing did both.
Boeing 787 Dreamliner
On paper, customers were impressed; by the end of 2007, the 787 was the hottest-selling jet in history. But when the first one took wing for its first test flight in December, it was the most delayed commercial jet in the firm’s history–28 months behind schedule. The 787 had become bogged down in a saga of parts shortages, subcontractor failures, and weaknesses in crucial composite structures, requiring retrofits and redesigns.
Where’s the Duct Tape?
So what went wrong? For starters, the company lost track of certain details–namely, fasteners. Building a single Boeing 777 requires 2.7 million titanium, aluminum, and stainless-steel bolts, rivets, and other fasteners peculiar to airframe manufacture. And the 787 would need more fasteners made of the titanium alloys that are least susceptible to corrosion when in contact with carbon composites.
As the 787 project geared up, the industry was already in the midst of a fastener shortage. But Boeing’s extensive outsourcing strategy compounded this problem. The back of the fuselage was made by Vought Aircraft Industries in South Carolina; a middle chunk by Alenia Aeronautica in Italy; the nose by Spirit AeroSystems in Kansas. In Japan, Mitsubishi, Fuji, and Kawasaki Heavy Industries built the wing structures–and Kawasaki built yet another piece of fuselage.
Suppliers were ordering fasteners in different ways and on different schedules; as a result, the fastener manufacturers found it difficult to make coherent production plans, according to a recent case study by the University of Michigan’s Ross School of Business that drew on Boeing reports and interviews with company employees. This bogged down manufacturing. “Boeing was caught off guard,” says Ravi Anupindi, a professor of operations management at Ross. “By the time they knew about it, it was at a crisis stage.” To solve the problem, the company wound up taking over the ordering of all fasteners.
Next, problems arose with the composite structures. In March 2008, Boeing said that parts of the center wing box, built by Fuji Heavy Industries, had unexpectedly buckled during stress testing. This caused a six-month setback as Boeing added aluminum reinforcements to the boxes and changed the designs of future ones. Then, days before a planned first flight in June 2009, the company discovered that composite “stringers” joining the main wing structure to the center wing box–the most severely stressed connections on any plane–had delaminated in testing. Boeing had to remake the wing-body connections, adding, among other things, 34 new titanium fittings.
The full story behind these issues has not been revealed, and Boeing provided no interviews for this story. “It’s hard to tell where a lack of oversight by Boeing ends and a bad contractor performance begins,” says Richard Aboulafia, vice president of the Teal Group, a think tank and consultant to the civil and military aviation industries. “Getting other people to build things for them worked well for Boeing, decade after decade. So with the 787, it was just faith. It might have worked with traditional designs, but with composites and new techniques, it was guaranteed to be a disaster.”
The problems and delays contributed to the $3.5 billion in charges the company took last year. And Boeing found it necessary to buy the South Carolina plant of Vought, which had fallen behind on its fuselage work, and to build a second fuselage assembly line. The company recognized belatedly that its outsourcing had gone too far, says Morris Cohen, a professor of manufacturing and logistics at the Wharton School at the University of Pennsylvania. “We see more and more outsourcing of manufacturing, globally, in high-tech industries,” Cohen says. “We are discovering that as you move down that path, the challenges are not trivial. I don’t think companies have paid enough attention to how to manage supply chains from a strategic perspective.”
The clearest consequence, besides the costly delays, is that the first few 787s will probably be heavier than the targeted 108 tons. Nevertheless, the first 787s are scheduled to be delivered this year. And as long as the plane ultimately performs as advertised, the delays may not harm Boeing’s long-term reputation. “Boeing can very easily redeem itself by producing a product that the market wants,” says Aboulafia. Airlines and passengers, he adds, “remember the product, not how it was executed.”
Meanwhile, the Airbus A380 program has faced its own manufacturing glitches. More significant, the plane has found relatively few buyers; with a capacity between 525 and 853 passengers, it is simply too big to make sense for many airlines. Whereas Boeing has logged more than 800 orders for the 787, only about 200 have come in for the A380, earning the aircraft–whose maker is headquartered in Toulouse, France–the nickname “Toulouse Goose,” after Howard Hughes’s infamous “Spruce Goose,” the wood-framed World War II-era behemoth that never made it beyond a single prototype.
Even if the 787 passes through all the turbulence, Boeing can’t rest for long. Airbus now has a 787 competitor waiting in the wings: the midsize A350, also 50 percent composite. About 500 orders have come in, and Airbus says it’s on track to start delivering the aircraft in 2013. So in a sense, the 787 program seems likely to pay off one way or another. If it provided a hard lesson about the limits of outsourcing and the risks of innovating with composite materials, the bold design also gave Boeing a much-needed head start on the next phase of innovation in commercial aviation.
David Talbot is Technology Review’s chief correspondent.
This new data poisoning tool lets artists fight back against generative AI
The tool, called Nightshade, messes up training data in ways that could cause serious damage to image-generating AI models.
The Biggest Questions: What is death?
New neuroscience is challenging our understanding of the dying process—bringing opportunities for the living.
Rogue superintelligence and merging with machines: Inside the mind of OpenAI’s chief scientist
An exclusive conversation with Ilya Sutskever on his fears for the future of AI and why they’ve made him change the focus of his life’s work.
How to fix the internet
If we want online discourse to improve, we need to move beyond the big platforms.
Get the latest updates from
MIT Technology Review
Discover special offers, top stories, upcoming events, and more.