Harper’s Plastic Journalism
This article in Harper’s about the project to create the Boeing 787 Dreamliner was written by a human being who has never done anything to me. Still, I feel entitled to crap all over it.
I actually do have some expertise as a third-party observer of Boeings creation of the 787 Dreamliner. The fate of the aircraft in many ways inspired my dissertation, and I’ve followed every twist and turn since before it was given the name “Dreamliner”.
Andrew Cockburn’s thesis is that the defense-contracting side of Boeing’s business infected the commercial-aviation side and that this negatively affected “the fundamental design concepts built into this revolutionary airliner”. This is a defensible claim, but it’s one that requires extraordinary evidence, of which there is none.
Here are the ideas presented:
- Sometime after the merger with McDonnell Douglas, Henry Stonecipher (of McDonnell Douglas) was made CEO of Boeing.
Changing a culture merely by switching CEOs is difficult. Stonecipher became CEO in December of 2003 and left in March of 2005. It seems unlikely that his presence alone could have the dramatic effects claimed.
- Defense contractors use outsourcing extensively.
Mr. Cockburn suggests that outsourcing occurs to that production is distributed across multiple U.S. states so that all Senators have something to lose. Boeing’s use of outsourcing, however, is fully explained by the desire to reduce its capital investment, the desire to limit the scope of its own engineering responsibilities, the need to access technologies held by other companies, and a need to reduce production costs. These explanations are preferable not only because they are the stated reasons provided by Boeing to its shareholders under penalty of jail time, but also because they explain why a whole bunch of manufacturers other than Boeing outsource. We do not need to resort to unlikely explanations for outsourcing when obvious ones are readily available.
- Boeing embraced risky technologies.
Boeing did implement multiple novel technologies in the Dreamliner, but there is no reason to suggest that this is an artifact of the defense industry. Rather, it is an artifact of the aerospace industry. When an aerospace company creates a new platform for the next 20-40 years, it’s going to involve the use of risky technologies that need to be worked out along the way. That’s part of why only two companies are left in the segment: launching a new platform requires a company to stake its future on its ability to work out whatever they have proposed. If Boeing could have gotten away with producing 737s forever and ever, they would have. But they couldn’t, so they tried to create the Dreamliner. And that meant accepting risk.
I am not a fan of the military-industrial complex. Part of me wishes that I Cockburn’s claims were true so that I could cite this case as an example of a once-great company done in by too much entanglement with the government. But that just isn’t what happened.
I used to Fisk Andrew Cockburn mercilessly, but his daughter is pretty hot (Olivia Wilde).
My issue with Boeing on the Dreamliner is that the lithium batteries they selected would have questionable safety (ie. non-perfect) in a consumer product, much less an airliner where the engineers keep pushing back the decimal places on in-flight mechanical problems. Boeing says the new batteries allow the same performance as NiCd’s with a 30% weight reduction, and I’d argue that the weight reduction (perhaps two passenger seats’ worth) probably doesn’t outweigh the potential risks, damage to Boeing’s reputation, the several-month’s grounding, and overall doubts about the 787’s safety.
I think they should have taken a more conservative approach and used an older, more proven battery technology to support the electrical loads, absorbing the weight penalty, and then introduced lighter battery technologies in later models of the Dreamliner instead of having new, relatively unproven battery system in the critical development path. At the very least, they should have an alternate system (with weight penalty) waiting in the wings as a fall-back option.
However, I’ve been told by one Boeing engineer that other battery technologies simply couldn’t have worked, an assertion I doubt because volts and amps are volts and amps. Maybe they couldn’t have met the performance figures their salesmen were using in the marketplace, but I imagine the 787 would still outperform older airliner’s even if it used lead-acid truck batteries.
As an aside, I think it wasn’t until the early or mid 90’s that new airliners caught up to the fuel efficiency per seat that piston-engine airliners had established by the 1950’s. We traded speed for fuel efficiency when we switched to jets, and gasoline engines weren’t even very efficient anyway.Report
Retrospectively, I, you, Andrew, and probably Boeing agree that they should have gone a different direction with the batteries. Then again, that’s retrospectively. There was a big long list of things that people could have pointed to and said “this seems like it could go wrong”. If they avoided all of those things, they’d have a 737.
I think it’s unlikely that lithium batteries ought to *never* be used on aircraft. They certainly need extra care, and it seems the level of care needed was higher than what was put in.Report
I’m a little hazy on the specifics — it’s been awhile — but as I recall, the Dreamliner is a really good example of how outsourcing can go wrong. Well, not exactly. More of a case in how outsourcing isn’t as simple to price as it appears.
Outsourcing can cut labor costs, but can also increase other costs — and Boeing found the latter, in this case, was far, far higher than the former.
I wouldn’t blame a “CEO culture” or “defense contracting” — not when a simple desire to cut bottom-line costs is sufficient. Outsourcing often shows immediate savings, which is sufficient right there for a lot of people to do it.
The costs come much further down the line. Sometimes it makes sense to outsource, sometimes it doesn’t.
Had I been running Boeing, I’d have hesitated to outsource critical components or key skills, nor would I have outsourced integration — at least initially. It just seems…inherently risky for low returns to outsource such things.Report
> Had I been running Boeing, I’d have hesitated to outsource critical components or key skills
I want to stress that it is *really* hard to know what we would have done in that position with the information that was available at the time rather than what we now know to be true. (That doesn’t mean that you’re wrong though.)Report
Oh the tales I could tell of the where’s & why’s of a lot of the decisions you are speculating on (I worked on the 787 as a Boeing engineer).
Sadly, my wife still works for Boeing, and me airing the dirty laundry is not wise…Report
I have similar tales about a lot of aerospace firms. 🙂 But we work with a variety of them, so I don’t tell them in public.
I’ve been ground central for a few high-class screwups, and the hind-sight reviews are often fun. Thankfully I wasn’t involved directly, so much as just impacted. It’s never fun to have that spotlight on you.
Now offshoring I know far more about, although I’ve been lucky never to lose my job to it. I’m sure some places it saves money, but I’ve yet to see it work long-term for software engineering. Outsourcing is often the same — sure, in theory good management and design and blahy-blahy means everyone can work from anywhere, but in practice it rarely works.
If nothing else, having the customers and the designers on different continents seems to really screw the process.Report
Let’s just say that the biggest defect in the outsourcing was more a matter of contract wording than anything else.Report
Well, perhaps you can answer a question I have. How come the Emergency Transponder Beacon System doesn’t have the self-diagnostic programming to broadcast “Holy F***! I’m on fire and I could torch the whole galley!”
Seems like they would’ve built that in. ^_^Report
I’m sorry, Boeing is NOT a software company, and you will get in trouble for even implying that they write software, any software, of any kind, ever.
(I spent 18 months not writing software, and one of my friends continues to not write software, etc.)Report
I’m sorry, Boeing is NOT a software company…
Sigh… beginning around 1981, in a different industry, I was the voice making presentations to management that said, “It’s a software world. Your future projects will succeed and fail, be delivered on time or late, largely on the basis of the software.” I lived through the limiting factor on the introduction of new services being what the billing software could handle. UPS succeeds or fails based on how well its scheduling software works. I sat through multiple post-mortems on government programs that failed because the software failed. Yeah, getting the battery chemistry right is important — but not as important as the software in the charging system that keeps everything operating within safe parameters. I love the example from a recent GM hybrid transmission — two-thirds of all the engineering hours were for writing and debugging the software running on the dedicated processor.Report
I just saw a new Wired article on the Airbus A350 composite airliner, which is a 350 to 400 seat airliner that will directly compete with the Boeing 777.
For the electrical system they stayed with NiCd’s.Report
For all the excited babbling about High Technology in Silicon Valley, Silicon Valley largely started with Lockheed doing the Polaris missile in Sunnyvale. Pretty much everyone in the neighbourhood where Steve Jobs and Steve Wozniak lived worked for Lockheed.
The Highest Tech I can think of is the logistics of the grocery store. Getting fresh fruit to market from halfway around the planet is just fascinating. That, and NASCAR. Which is sorta odd, because I don’t like NASCAR as a sport. I just like the tech.
George and I kinda agree on the battery issue. NiCd is a more reliable technology but there doesn’t seem to be a good reason why Li-ion can’t be done safely. It’s all in the form factor: putting all that lithium in one pack isn’t smart. It could and should be distributed better, attenuating fire risk.Report
Changing a culture merely by switching CEOs is difficult. Stonecipher became CEO in December of 2003 and left in March of 2005. It seems unlikely that his presence alone could have the dramatic effects claimed.
Assuming that a CEO is coming into an entrenched corporate culture (i.e. General Motors), it is very difficult for a CEO to make any kind of change in a very short period of time. However, given that the two firms merged in 1997 and Stonecipher took the CEO position in 2003, it does suggest the possibility that the underlying corporate culture changed enough that Stonecipher could emerge as the CEO. These things don’t happen in a vacuum nor without some kind of internal power struggle. I haven’t done any research on this, but I wouldn’t be surprised to hear that were people on the old Boeing side that thought it should have gone to one of their own so to speak.
Great post.Report
One minor quibble: “These explanations are preferable not only because they are the stated reasons provided by Boeing to its shareholders under penalty of jail time, but also because they explain why a whole bunch of manufacturers other than Boeing outsource.”
I would rate it as more likely that a commenter here will be hit by lightening than the CEO or other high officer of a powerful megacorp is indicted, much less convicted, much less does time, for lies of this nature.Report
@Dave, the plan had been for Stonecipher to fade off into the distance. He was called back to become CEO because the prior guy got caught up in a scandal or two. (This doesn’t directly contradict what you said, but I thought the context might be useful.)
(Incidentally, Stonecipher himself left because of a scandal.)
@Barry,
That’s a fair quibble.Report
MRS,
yeah, The stories… I know a guy who got called in when the shit hit the fan on the Dreamliner.Report
When I see news that a big corporation has a new CEO, my assumption is that the new CEO isn’t coming in alone, but rather that she has a whole team of people she’s assembled coming in with her. Heads are rolling, new broom sweeps clean, and so on. A CEO on her own may not be able to do much to change a corporate culture. A CEO with a loyal team that shares her vision and has invested in the company’s success can. A CEO without such a team will probably be a weak leader.Report