The Boeing 787 Dreamliner lithium-ion (Li-ion) battery fiasco that I wrote about here several weeks ago continues to make news. A few examples:
Airlines that had taken delivery of 787s before the grounding have canceled thousands of flights and are canceling thousands more. All Nippon Airways alone has canceled more than 3,600 flights through to the end of June. Other airlines are putting off some new routes (such as United’s nonstop from San Francisco to Paris) because they need those planes as replacements for their 787s.
Airbus, which also had planned on using Li-ion batteries, has changed its mind and will stick with nickel cadmium batteries (a striking example of benefiting from not being first to market).
Boeing has a parking problem with its growing backlog of assembled but undeliverable 787s. The company has continued to produce additional 787s as if it can resume deliveries to customers tomorrow, but there’s no more parking space at its assembly plants, and the planes can’t be sent to other airfields until the FAA provides waivers.
It seems that no one even knows why the batteries failed—or whether the batteries are even at fault. Japanese air safety officials say there’s insufficient evidence to refute their hypothesis that surges in electrical power outside the batteries, not the batteries themselves, may be responsible for the overheating. (The batteries were developed and produced by a Japanese company, GS Yuasa.) Boeing can’t expect the aviation authorities to lift the flight ban if experts can’t even agree on why the batteries are misbehaving.
All Nippon revealed that it had to replace 10 batteries between May and December last year. Those defects preceded the fire and smoke issues that resulted in the 787 grounding. Those problems were not reported to the National Transportation Safety Board (NTSB).
On March 11, the FAA approved Boeing’s plan to test fixes for the battery problem. Boeing’s new battery design includes better protection in case a battery overheats. The FAA could still demand changes if problems develop in laboratory and flight tests. Boeing hopes to begin fitting the redesigned batteries in the grounded 787 fleet by mid- to late April and resume commercial flights quickly after that, but government officials are not sure the process will move that fast.
Perhaps the most consequential item of all concerns restrictions to flight range. A major selling feature of the 787 was its ability to reach almost any airport in the world without stops. For safety in the event of problems such as engine failure, most jetliners are required to stay within three hours flight time from the nearest airport. The 787 was expected to be approved for 5.5 hours. Three hours separation from the nearest airport presents no problems for most flights in the northern hemisphere, 5.5 hours makes a big difference in fuel efficiency for travel to areas such as Australia. Until the FAA is convinced that enclosures for the 787’s batteries will prevent fire and smoke from escaping to the rest of the plane, there is low probability that Boeing will receive approval for the greater separation from airports.
- The NTSB has announced that it “will be holding both a forum and a hearing in April to provide additional information to advance the investigation” into causes of the battery failures and Boeing’s certification procedures that led to acceptance of the Li-ion technology. So Boeing now must satisfy two U.S. government agencies—in addition to similar agencies in other countries—that the 787 is safe.
And those are by no means all of the problems afflicting Boeing.
Several readers posted comments on my original blog, arguing the battery is a minor blip that Boeing will correct quickly. My impression is that the battery is becoming aviation’s equivalent of the Ford Pinto, a 1970s model killed off by reports of faulty gas tank design. Unlike Ford with the Pinto, however, Boeing can’t take the 787 off the market. The accumulation of bad news will probably reduce airline demand for Dreamliners in the short term, pushing out the time when Boeing begins to make money on the plane.
Interestingly, while the 787’s Li-ion battery has not been used in commercial planes before, the Boeing batteries are not cutting edge technology by today’s standards. Rather, the battery uses old (in battery development terms) flawed technology more prone to thermal runaway than newer forms of Li-ion. In 2010, the FAA tested several types of Li-ion technologies and found that the type chosen by Boeing was the most flammable. IEEE has a good technical explanation of the 787 battery technology and a photo of the charred 787 Li-ion battery.
Cessna tried Li-ion batteries but reverted to nickel-cadmium batteries after a 2011 incident in which a Li-ion battery began smoking during charging. Investigators found that recharging a severely drained Li-ion creates substantial heat. Boeing says systems on the 787 do not allow recharging of greatly depleted batteries; the batteries must be sent back to the manufacturer for reconditioning.
Several comments on my previous blog argued that Boeing is a great company and the battery issue is just a tiny hiccup. There’s no question that Boeing is, indeed, a magnificent company that has transformed air travel. But I don’t consider the battery issue trivial; just the bullet points above translate into huge costs for Boeing and its customers. And, since the number of 787 flight hours is small and the plane’s construction breaks a vast amount of new and complex engineering ground, it’s possible that other nasty surprises remain to be discovered. As Donald Rumsfeld would say, there are unknown unknowns.
Although the battery story has attracted the most attention, I believe the focus should be on Boeing’s supply chain strategy. After all, the battery mishap seems to have supply chain mismanagement as the root cause.
As I noted in my last blog, Boeing abandoned its traditional approach of designing in-house rather than relying on third parties. Boeing has used suppliers for actual fabrication of parts in the past, but the design of those parts was always performed by Boeing engineers; suppliers were subject to rigorous Boeing oversight. For the 787, Boeing broke the mold.
According to Forbes, 60 percent of the design and production of the Dreamliner was performed by suppliers rather than Boeing itself. Those suppliers were responsible for the entire cost of development and production for their particular hardware. They get paid only as planes are delivered to airlines. (If Boeing is hurting because of the grounding order, imagine what life is like at some of its less-flush suppliers.) Boeing’s risk-averse management believed that this arrangement would provide them with better parts faster and at lower cost. Instead, they lost control. Suppliers ended up controlling Boeing, rather than Boeing controlling suppliers.
Some of the consequences are well-known. Delivery of the first 787 was delayed by more than three years. Not all the pieces fit together; apparently a gap of several inches was found where the cockpit joined the cabin. Boeing anticipated that the 787 would be profitable by 2015, but a UBS aerospace analyst says that will not be possible until 2021 unless Boeing manages to bring costs down quickly and substantially. Can Boeing achieve meaningful cost reductions when so much of the hardware is controlled by the supply chain?
Defenders of Boeing’s strategy argue that it’s not much different from the way the automotive industry has operated for the past couple of decades. Superficially, that seems to make sense. These days, automakers are primarily assemblers of components designed and built by third-party suppliers. But the argument breaks down on closer examination because an aircraft—especially a large passenger jet—is not like an automobile in several important ways:
Cars and trucks sell in the tens, sometimes hundreds, of thousands every year. Boeing will be ecstatic if it sells more than 1,200 Dreamliners over the next couple of decades. This matters because volume manufacturing is easier to control than batch production.
Automotive manufacturers have 100 years of experience for selecting and managing suppliers. Boeing doesn’t have such deep experience. Nor, for that matter, do other aircraft manufacturers.
Most automotive suppliers have been making the same kinds of parts forever. They understand the development cycle. FMEA and 8D are close to religions in the automotive world. In the case of the 787’s batteries, Boeing seems to have known since at least 2010 that the batteries were prone to overheating and other failures, yet nothing was done. In the auto world, such disregard for signs of trouble is inconceivable. New suppliers get managed very aggressively by the automakers’ design engineers and supplier quality personnel until they mature.
Risk is exponentially greater in aircraft failures than vehicle malfunctions. New models of cars and trucks are rarely free of problems upon launch, but the defects result in inconvenienced customers rather than loss of life. At worst, a vehicle failure rarely costs more than a handful of lives. More often, the problems involve failure to start or keep running rather than loss of control at high speed. It’s a different story when an engine or electrical system fails at 35,000 feet.
- The development cycle is much shorter for vehicles than aircraft. Most new automobiles are in showrooms within three years from the first day of development. Planning for the 787 began more than a decade ago. Every new product, whether a car or a plane, requires some guesswork about whether unproven innovations will actually be ready for the real world in time for the new vehicle launch. Boeing and its suppliers had to bet that technologies little more than concepts in 2002 would be ready for use by 2008, the expected launch date. As it turned out, of course, much new technology wasn’t ready for prime time in 2008, and some of it (like the Li-ion batteries) still had bugs sufficiently serious to push the 787 launch back by three years.
So, here’s my take on how Boeing handled expansion of its supply chain: The company took on too much risk at one time. In seeking to avoid major development costs, it actually ended up more exposed. Ultimately, the gamble may—probably will—pay off. But, so far, the return on investment amounts to years of launch delay and the most embarrassing recall in aviation history.
Supply chain true believers will likely disagree with my assessment. What do you think?
Editor’s note: Before “Shipulski on Design,” “Leading Lean,” and “Uncommon Sense,” there was ASSEMBLY magazine’s longest running and most controversial back-of-the-book column, “Unconventional Wisdom” by Jim Smith. A nationally known expert on electronics assembly, Smith never hesitates to question the sacred cows of manufacturing and economics. You can read more from him at his “Science of Soldering” blog.