The Myth Of Old Boeing

A door blew out of a brand-new Boeing 737 MAX 9, and fingers are being pointed at the company’s apparent lapse in quality inspection, and at Boeing’s cost-cutting and focus on shareholder value as a root cause. Not without good reason: suppliers like Spirit, spun off in 2005 and 50% dependent on the 737 program alone, have been squeezed until the pips squeak by the primes, and once the product arrives at Boeing, the pressure is on to maintain the pace of production. That’s not a healthy atmosphere.

But then people start telling a story that has heroes and villains, and while this is comforting, such stories are seldom complete.

The villains in this tale are the invasive species from St. Louis, who managed to seize control at Boeing even though it was their company, McDonnell Douglas (MDC), that had been acquired. Short history: MDC had let its commercial product line obsolesce, losing second place to Airbus during the 1980s. It suffered from erratic management in the early 1990s and then, to teh surprise of many, was shut out of the Joint Strike Fighter competition – then the entire future of US combat aircraft – by Boeing, which hadn’t built a production fighter since the 1930s.

Weeks after the loss, it was announced that Boeing would buy MDC.

The deal must have been secretly in the works for months before that, contingent on the JSF result, so one person at MDC who had not been 100% confident in the JSF outcome was the deal’s almost-certain architect. Harry Stonecipher had been brought into MDC as second-in-command to chair John McDonnell in 1994. McDonnell, the founder’s grandson, had talked airily of a future collaboration with Airbus – to the complete surprise of, errm, Airbus. Stonecipher realized that MDC’s commercial side was history, and knew that the Pentagon was determined to reduce the number of combat aircraft primes, with the 3000-aircraft JSF as carrot and stick. If Macs did not win JSF, its value to Boeing would never be higher again. Part of the deal was that Stonecipher himself came to Boeing as president under CEO Phil Condit.

Trained at General Electric under Jack Welch, Stonecipher did indeed instill a GE culture of cost-awareness at Boeing. It was the start of the shareholder-value focus, the start of outsourcing, the start of kneeling on suppliers.

The myth is that pre-merger Boeing was some kind of Platonic ideal of an engineering company, where excellence in product design, manufacture, and support was the way to commercial success.

Boeing had done well for itself, but it had also enjoyed good luck. Its US rivals, Lockheed and McDonnell Douglas, had dueled over what a limited market for the L-1011 TriStar and DC-10. The McDonnell bosses, who had been strong-armed into taking over the ailing Douglas in 1967, never invested in an all-new airplane once the DC-10 was in service. Then, widespread deregulation of air fares boosted international travel – good news for Boeing, because until then the 747 was too big for many routes.

Boeing guessed wrong when it came to shorter-haul services, betting heavily on the 757 – but was saved by a small team who had been given some spare change to do something with the 737, the runt of the Boeing litter. But then, increasingly focused on beating Airbus, Boeing blitzed the world with marketing and promotion for the prop-fan-powered 7J7, intended to annihilate Airbus’ new A320. It was a bust, technically and economically unfeasible.

Boeing’s next new launch, the 777, was a big success with the customers – but overran its development costs on a grand scale. Fortunately it entered service in a recovering market after the early 90s slump.

By 1997, Boeing has preparing to deliver the latest 737 Next Generation by the bucket-load. It failed. Deliveries were late, and penalties were piling up. The cost of production was getting perilously near the sales price. Arriving in Seattle in August 1997 as the deal closed, Stonecipher looked at what was going on.

Boeing was really, really bad at building airplanes.

The core problem was “configuration control” which in essence was the path from the drawings that defined, for example, a Southwest Airlines 737-700, to the procurement of the numbered parts on those drawings, and delivering those parts to the correct point on the assembly line. Boeing’s system was antiquated, as Stonecipher told me in August 1998. “I was very surprised to find that system in place when I arrived.” Stonecipher had never seen anything like it since he started at Allison in 1955, and called a friend who was familiar with production systems. “He was appalled. He said: `It’s not 1955, its World War 2’.”

“I think that everyone believed that it was another problem, one that we could power through,” Stonecipher said. “The attitude was that `We’re at our best when we get in these situations’.” Stonecipher, the outsider, saw the problem differently:  Boeing’s strength in the market had allowed it to survive and prosper despite the fact that it was inefficient, and the company’s failure to carry out its planned ramp-up in 1997 deliveries showed that it could not live that way for ever. “Even if we had been able to execute our plans, and had we not missed a delivery, the underlying cost was eating away at us.”

To put airplanes together and track and identify parts, Boeing relied on an arcane doctrine called “effectivity”. The drawings defined the airplane. The problem was that the doctrine had indeed been developed in WW2 – when every airplane in a given production block (for example, a B-17G-35-BO was in the 35th B-17G Block) was identical. The system had been racked and tormented and bodged over decades to cope with the fact that every airline insisted on different optional equipment and every aircraft was built with different weight options.

The bill of materials – the list of part numbers attached to each drawing – began to grow. And grow. By 1997 the 737 landing-gear drawing carried 464 pages to determine which optional parts were attached to each aircraft.

Could things get worse than this? Unfortunately, they could. Over time, Boeing’s engineering department developed its own computerized system to handle the bill of materials. So did the manufacturing department, and the product support department, and… ultimately, the bill of materials metastasized into 14 systems, with data hand-entered from one to another. (This detail was all explained to me in a remarkably frank interview by vp Bob Hammer. I realized why he was being so direct when I heard he retired two months later. He wanted to unburden himself.)

4,000 parts in Wichita were scrapped because an engineering change did not make it to the production line in time. When Boeing changed the paint on the bilge sections of the 737 and 757, in 1991, the job took 5,000 hours of engineering effort. Changing drawings to reflect the changes consumed 105,000 hours.

The only reason the system didn’t crash long before it did was that Boeing had a core of veterans who understood it better than most, had worked with it man and boy and could make it function. But they were mostly not high-level people, and when Boeing offered an early-exit package in the 1991-92 slump, a lot of them took it.

It wasn’t that the problem was unrecognized – the effort that Hammer was running, called Define and Control Aircraft Configuration/Manufacturing Resource Management (DCAC/MRM), started in late 1993 and was expected to take 12-18 months – but that nobody had a clue how bad it was.

As Hammer’s team attempted to resolve the problem, they identified the company’s “skin and spar” shops as the first target for new systems, “because we thought they would be easier.” He recalled: “We asked how many part numbers they dealt with, and they said they had 500-550.” Hammer’s team asked them how many systems in the company used their part numbers:  their estimate was about a dozen. The skin and spar shops have 2,500 part numbers, which are used in 109 systems.

“You can’t talk to me for more than five minutes without talking about cost,” said Stonecipher. He toured Boeing’s myriad subassembly shops, and the experience had sometimes been frustrating. “You go into some shops and ask what something costs, and they can’t tell you within a row of apple trees.”

Stonecipher made Hammer’s project top priority. His methods were direct. A fervent apostle for the kind of high-technology, low-manhours production approaches practiced by McDonnell Douglas’ Phantom Works, Stonecipher told the manager of one Boeing plant to consult Jerry Ennis, leader of the Phantom Works, about high-speed machining technology. “A month ago I saw Jerry, and he hadn’t had a call,” Stonecipher related. The next time Stonecipher saw the manager at a meeting, he walked over to him and put both hands on his shoulders – and Stonecipher is not a small man.  “You didn’t call Jerry Ennis”, he said. “I thought he was going to melt, right down to his shoes,” Stonecipher recalled. “If these people had known me a little longer, that would have been a totally different conversation. When I suggest something, I expect someone to go do it – or tell me why we shouldn’t do it.”

So whatever happened later, pre-merger Boeing needed a shot of what St Louis had. Maybe the methods were extreme: later, I characterized Commercial Airplanes’ relationship with corporate as YOU’RE NOT MY REAL MOMMY AND I DON’T HAVE TO DO ANYTHING YOU SAY. It’s gone over the top – but then, it was the Seattle veterans, Condit and Alan Mulally, who gave us the Sonic Cruiser and defined a 787 program that, even in 2003, looked risky.