Is it time to revisit (or visit for the first time) some of the central challenges of developing and launching sustainable innovations? With the demise of Solyndra and Beacon Power still recent memories; with Ener1 entering bankruptcy; and the recent disclosures that EV makers Fisker Automotive and Tesla are troubled, it may be long overdue.
Most of the attention in cleantech goes to cutting edge science and cool new startups, to the bold ideas that are going to lead us painlessly into a low-carbon future, hopefully with the same rapture the iPhone brought. Yet in talking to companies for The Business of Innovating: Bringing Low-Carbon Solutions to Market, one of the biggest lessons I learned lay with the 'boots-on-the-ground' challenges in bringing sustainable innovations to market.
Bold new ideas are nice, but the real challenges lie in making them real. Sustainable innovations are by definition going after brownfield problems—attempting to displace existing practices that are widely-used, deeply-embedded, and (of course) inherently unsustainable. This is very different from going after greenfield problems (where there are few customers, if any, and few established practices to compete against). Launching a moonshot, or a search engine, or a search engine's attempt at a moonshot are greenfield problems.
When companies stumble bringing ideas to market, their stories become fodder for politicians and incumbents who want to disparage the very worth of new ideas. But Solyndra's fall is not an indictment of solar power and Tesla's troubles are not an indictment of electric vehicles. They're just evidence that good ideas, brilliant scientists, and dashing young entrepreneurs are not enough and we will waste a lot of time and money if we think they are.
"Faster, better, cheaper"
When I worked for IDEO, we played in a city softball league and our team shirts read, "Faster, better, cheaper. Pick any two." It was an inside joke. Regardless of the promises we made, in the consulting world clients could have their work done ahead of schedule, at the best quality, and under budget. They just couldn't have all three. It's one of the laws of consulting and probably all companies trying to innovate.
We've been able to ignore this law—spoiled as we've been by 50 years of IT revolutions. Moore's Law seemed to free us of the limits of faster-better-cheaper: Every year, products were getting to market faster, performing better, and costing less than the the year before. And the manufacturers of these products were able to make more of them, increase their reliability, and rake in a handsome profit too. But in many ways, the IT revolution took place in greenfields where markets started small, had low expectations, and were willing to pay anything for the latest improvements.
The current supermodels of innovation—the Apples, Googles, and Facebooks—are relatively immune to faster-better-cheaper. Apple has FoxConn to ensure they're products come out faster, better, and cheaper because in fact little changes in the manufacturing process. Google and Facebook launch new products by adding links to their website, and assigning a few more servers to the task. If its buggy, they just call it Beta.
In the real world, outside the Silicon Valley, most companies and industries must live within the constraints of faster-better-cheaper. When innovations are competing with widely-used, deeply-embedded, and inherently unsustainable competitors, the bar is set even higher. Bringing new sustainable innovations like solar, wind, electric cars, smart buildings, etc.. requires reaching manufacturing scales that equal incumbent technologies (faster); requires making products as reliable as incumbent technologies (better); and requires making them profitably in spite of the low margins of most entrenched competitors (cheaper).
For new ideas to have an appreciable impact, they have to come to market at the same scale as the incumbents they're trying to replace. Solar PV is more sustainable than coal but, represents only 0.7% of worldwide electricity capacity now and, by 2030, will reach only 1.5% (International Energy Outlook 2011; for a good graphic, see Small Gains). Automobiles and trucks, household appliances, industrial motors and manufacturing processes, power plants—these are all huge markets.
Take Tesla, which faced the scale challenge in introducing its electric Roadster. Unveiled in 2006, the first production car wasn't delivered until 2008, and Tesla struggled over the next two years to build out the supply chain (Tesla Production Slower Than Expected). By the end of 2012, Tesla reports, they expect to have sold a total of 2,500 Roadsters (by contrast, DeLorean sold approximately 9,000 cars in its single year in production).
This is no indictment of the electric vehicle. Toyota made and sold 18,000 Prius vehicles in 1998, the first year it was introduced, due mainly to Toyota’s previous experience and established capacity (they now annually sell over 3M cars worldwide). And for the new all-electrics in 2011, Nissan sold roughly 10,000 of the Leaf while GM sold just under 8,000 of its Volt in their first years.
Instead it's a recognition that unless a company already has most of its supply chain built out, getting to scale is really hard. Any new manufacturers going after large markets, whether automobiles, utility-scale energy, health, nutrition, agriculture, water treatment, faces the same challenges. They must not only develop a new product but, if the market embraces it, build out the supply chain and manufacturing capacity to meet that market demand. Designing a cool new electric car is like a dog chasing a bus—the only thing harder is what happens when you catch it.
Reliability refers to producing new goods and services that perform as promised. In many cases, that doesn't mean they work flawlessly—the early days of the computer industry were characterized by frequent system crashes, hard-disk failures, and other reliability issues. So too were the early days of the automobile industry. As markets mature, however, those problems become intolerable and in many cases even regulated against (e.g., the automobile, food, and energy industries). Martin Daum, CEO of Daimler Trucks North America, decribed the challenges of designing the new and radically more efficient Freightliner Cascadia long-haul trucks:
You have to make the technology mass-production capable, and then you have to test and make sure it is absolutely 100 percent reliable…. The customer in our industry expects nearly 100 percent reliability.
Reliability is easy, if you're willing to give up scale or profitability. Witness Solyndra, which despite their reliability problems, tried to move from lab-scale production to mass production. The company had developed a novel approach to solar panels and if it they could have delayed scaling up, they might have survived. But they still had a buggy product when they felt their investors (and a $535 million loan guarantee from the DOE) breathing down their necks to scale up—and so were caught trying to fix reliability while scaling up (and running out of money). Tesla and Fisker, too, experienced quality problems in their rush to meet promised production volumes and schedules. For Tesla, Early transmissions failed in the field and a faulty rear hub required a safety recall of the first 345 units produced. For Fisker, a flawed battery assembly design caused a safety recall and the test car for Consumer Reports shut down during the testing
The problem isn't limited to startups either. GM's Volt was hit by two embarrassing problems that emerged in the field: a crash-induced risk of battery fires (which prompted a voluntary recall of 8,000 vehicles) and melting electrical plugs.
And it's not just reliability of the product. Wal-Mart’s decision to purchase LED lighting for their parking lots hinged not on the energy savings, but on the long-term savings in the maintenance costs of replacing bulbs (LED lights have a longer theoretical life than traditional metal halide lamps). Wal-Mart chose General Electric rather than smaller startups because they wanted to be sure the company would be around as long as the 12-year warranty... matching the LED’s useful life (see MIT Technology Review's LEDs Are Getting Ready for the Spotlight).
Finally, companies must ultimately be profitable. Scaling production and managing reliability is costly. Wendy Graham, an executive at Air Products and Chemicals, Inc., which provides industrial gases and other process technologies for power generation, emphasized,
“There is a real scale difference between typical innovation and low-carbon innovation, especially the need to establish reliability and incentives for low-carbon solutions before energy industry customers will adopt them. Utility customers are interested in large-scale projects that have been demonstrated, and [that effort] can take hundreds of millions of dollars.”
Scaling too quickly or outpacing reliability brings unexpected material, operational, and warranty costs. It's especially difficult when introducing new products into mature markets where incumbents, enabled by already long histories and established supply chains, enjoy reliability and low costs.
Clipper Wind, Sinovel, and Suzlon Energy all faced unexpected quality problems stemming from wide-scale deployment of new wind turbine technologies and designs. In 2008, Suzlon Energy had to “strengthen or replace 1,251 turbine blades -- almost the entire number it has sold to date in the U.S. -- after cracks were found on more than 60 blades on turbines” installed and in use. The rapid deployment of wind turbines led to problems that proved extremely costly to turbine manufacturers—almost fatal.
So how do you deal with the faster-better-cheaper law—especially when introducing sustainable innovations? There's no easy answer except to recognize most innovations don't follow Moore's law.
For one thing, pacing is critical. Bob Swanson, founding CEO of Genentech, was renowned for his "strategic patience," his recognition that biotech had to proceed at its own pace, regardless of the urgency of the market or investors. To Swanson, neglecting that pace meant risking worse mistakes in reliability or costs.
Revolution Foods is pursuing the same strategic patience as it attempts to change the entrenched $7 billion market for K-12 school lunches (currently dominated by international food businesses serving frozen, processed foods). Their mission: the bring healthy, fresh, and organic food to students across the country. In spring of 2009, they prepared 14,000 lunches a day, now that number has grown to almost 70,000 lunches and across California, Colorado, Delaware, Maryland, New Jersey, New York, Texas, Virginia and Washington, D.C. The co-founders recognized that the only thing worse than not growing was growing badly.
It's one thing to experience dropped calls on your new iPhone when (1) it's not the end of the world when it does, (2) this thing allows you to do other things you never dreamed of, and (3) it's going to be obsolete in two years anyway. It's something wholly different when your livelihood (or health) depends on it, when it's doing the same job someone else is already doing reliably, or when its useful life is measured in decades (10 years for a car or truck, 40-60 years for a wind turbine or solar plant).
That means entrepreneurs and corporate managers introducing sustainable innovations must develop the strategic patience to recognize the proper balance between all three—when to loosen the demands on one or the other in order to stay in business. As consultants, it was always easier to let the budget slip—that mistake was forgotten if the work was good and things made it to COMDEX on time. For sustainable innovations, on the other hand, I suspect the better strategy is to be patient for scale.