6 Comments

Summary:

The lack of greentech hardware success stories points to a potentially missing piece of the puzzle: a Samsung-style intense focus on manufacturing.

MiaSole_0876[1]

Back in early 2000, I spent four years working in Samsung’s LCD division, where giant sized TVs were pumped out of their football field-sized manufacturing lines 24/7. Manufacturing at Samsung always had a “war-like” kind of atmosphere, where entire fabs would spring up from scratch in six months and every new process would be ramped up to high manufacturing yields in a short amount of time.

I imagine a similar focus on high volume manufacturing at other successful hardware companies like Intel and First Solar, that make millions of units per month. This is no ordinary feat and is achieved only by a disciplined approach, operational rigor and a clear vision that “he who controls manufacturing eventually controls the market.”

Japan auto industry factorySilicon Valley is the mecca of innovation worldwide, which makes the recent lack of successful greentech hardware startups difficult to comprehend. While there are a variety of reasons for this issue, one thing that might be missing from the greentech ecosystem is this Samsung-style intense focus on manufacturing.

Innovation is good, but scaled up manufacturing matters too.

Silicon Valley startups producing solar panels, batteries, energy efficient windows and other volume hardware products will have to achieve manufacturing at the scale of tens of thousands of defect-free units per month to make a dent in the marketplace. Having a superior product specification might not be enough if the companies are unable to produce the goods cost-effectively in mass numbers. In fact, their very ability, or inability, to do so might end up being the differentiator between survival and failure.

Technology transfer from R&D to manufacturing requires a different crew geared toward an operations-driven mindset and a brute force methodology that can quickly deliver a stable process. The 24/7 lifestyle that is required to deliver, say an 80 percent up-time, is painstaking, time consuming and requires dedication and commitment. At Samsung, the technology transfer engineers were treated like celebrities and provided with five star accommodations at the factory site. Special bonuses and incentives were tied to the speed of progress. One might say that’s an unfair comparison, but the reality is that Silicon Valley is competing against several such cleantech companies from all over Asia. Unless the startups have extremely revolutionary products that can tolerate lower manufacturing yields, there is a need to manufacture as efficiently, if not better, than the competition.

First Solar Malaysia plantHaving said all that, Silicon Valley was the birthplace of the “Intel Way” and has had a rich history of volume manufacturing. But an exodus of semiconductor fabs from the Valley to nearby states has taken the manufacturing focus out of the Valley and along with it, a lot of the manufacturing talent. This, in combination with the rise of Asia and the difficulty in obtaining financing for manufacturing has made it even more challenging for greentech manufacturing startups to scale up.

How to get around the manufacturing learning curve

Here are a few ways that greentech startups could help overcome issues with scaling up to volume manufacturing:

1). Obtain an experienced crew that can transfer R&D processes into a pilot line with an extreme focus on improving yields. Ensure that every employee understands that true success comes not from obtaining one-off records or champion results but from volume data obtained from processing hundreds of samples. At Samsung, no champion results were ever reported as an accomplishment in an upcoming line since those results are meaningless until proven at scale. Companies like Miasole in Silicon Valley took the right approach of involving Intel manufacturing experts for their solar panel factories and it would be wise for others to follow suit.

2). Manufacturing is a costly endeavor and gives the most bang for buck when it’s running 24/7. Hence it’s advisable to not wait for a perfect product but transfer an intermediate process and iron out kinks in the production process. A lower specification product with 90 percent yield might have better economics than having a higher performing product with a 50 percent yield, Thin Film Solar Underdog MiaSole Looks Ahead to New Plant, Solar Shinglesassuming the market will accept such a product for some time. There is no better example than First Solar who has continued to sell less than 12 percent efficient solar panels for a long time proving that there is room for a lower specification product if you make manufacturing extremely efficient.

3). Excellence in manufacturing depends on the company’s ability to expand and build future lines. For a Silicon Valley startup, a pilot line locally in the area might make sense to be close to the R&D team. But it might be wise to take the cue from other high tech companies and move manufacturing to nearby states given their lower costs of land and permitting, robust manufacturing ecosystem and availability of manufacturing talent. Several companies like Stion, Soladigm and others have already taken that approach, which is a good sign.

The importance of manufacturing to ensure a greentech hardware company’s success cannot be stressed enough. High-yield manufacturing allows a company to reduce the cost of its product and reinvest the profits into newer lines and more R&D. All of this requires a cultural change across the company.

The startups that truly understand the differences of personnel and resource requirements for manufacturing vis-à-vis R&D have a greater chance of success. Those that don’t might perish to the Asian giants who believe it or not, have figured out the art of volume manufacturing.

Kunal Girotra is a Director of Process Engineering at ThinSilicon, a solar cell R&D subsidiary of China Solar Power. Prior to that, he worked in various R&D and technology transfer roles at Samsung’s LCD division for advanced LCD and OLED displays. He can be reached at kunalgirotra@gmail.com.

Images courtesy of Miasole, Shutterstock/Mypokcik, First Solar.

  1. Great article. I think part of what made the Intel Way a success was the fact that early in their history, Intel’s products could initially be sold for a high margin, which gave them the headroom to tolerate screwups, excursions, poor yield, etc. before the competition caught up. By contrast, in cleantech startups, or at least in solar in particular, the margins got sucked out of the business so fast that no one really had a chance to figure things out before they ran out of money. (Some would say that Nanosolar and Miasole had more than enough time, but in their defense, they were developing a fundamentally new semiconductor system that had a lot more quirks than anyone could have predicted.)

    So what we have nowadays, with over half the world’s solar panels being made in China and vanishingly few made in the US and Europe, is a world in which the people who are best at manufacturing are largely disconnected from the people who are best at innovating. Something tells me that this cannot persist in the long run: Either Chinese companies will legitimately look to partner with Western companies/labs for new technologies, or more Western companies will pop up along the lines of SunPower, First Solar, Intel, etc. that take the track of innovation and development in the US/Europe, followed by eventual technology transfer to Asia. (Of course, it is also possible that China will figure out how to innovate as well as the West, but to me that would require a transformation of their society at large, something that is unlikely to happen anytime soon.)

    Anyway, that’s just my $0.02. Thanks for the thoughtful essay.

    Share
    1. Thanks Alex for your inputs. You have a good point on why Intel, First solar did well with manufacturing given they had time to iron out kinks before the competition caught up. And even though I feel for Nanosolar and Miasole given they are/were tackling really big and difficult problems, I do think the point to be conveyed is they probably innovated on too many fronts at once and expected that to scale without any problems. Experienced manufacturing people realize that you can’t implement all product innovations at scale at first try. You have to slowly ramp up with one or two innovations and slowly introduce next versions once the older concepts are up to scale. My guess is that Nanosolar and Miasole thought they could implement all their tricks at scale at once, but my assumption may be wrong since I haven’t spoken to any of their key manufacturing personnel. Thanks for reading my article though.

      Share
  2. Thanks for your article, nice analysis. actually, I would like to know more about “volume manufacturing” as you mentionned. Actually, what I know is that China is pushing strongly to reduce the CO2 emission and there will be a potential great market, however I heard much less in Europe about this industry whose key issue is actually R&D cost. Am I right?
    So when you talk about “volume manufacturing”, you should talk about also the sales and the potential market. Do you think that USA or Europe market coubld support it?

    Share
    1. I am sorry but I could not understand what your question is. The article assumes that the market and demand is ready. It also assumes the product specifications of the startups are superior to competition. Its a question of whether startups (esp based out of silicon valley) can execute manufacturing well enough to achieve cost effective products in mass numbers. The ability to do that will define their true success.

      Share
  3. Enjoyed thearticle. One potential pitfall that start ups might face while preparing for high volume manufacturing is predicting the rate at which the demand for their technology will grow. Case in point, electric vehicles where a lack of widespread acceptance (atleast till date) of the technology has forced the major players to scale back manufacturing and that strategy has impacted their suppliers, many of whom are startups. In many instances, these startups have been forced to establish their manufacturing bases in Asia which does not endear them to the political establishment or the common public in this country.

    Share
    1. Thanks Abhishek for reading the article and your comment. I completely agree with your take on the battery industry and the conundrum faced by upcoming startups in that space. Batteries for EV have stringent product expectations in terms of safety, reliability, weight, cost among other things. To satisfy all those constraints and produce them with a high yield in volumes is a challenge for any company, and more so for a startup where availability of finance is tough, and not to mention your point of uncertain demand. Its a tough road out there for startups in this space and other companies trying to solve real material science problems. Only companies with good financing and robust manufacturing will survive in my opinion!

      Share

Comments have been disabled for this post