And then Windows Became Intelligent …

Luke Pustejovsky
Mar 31, 2017 · 28 min read

Volume 1: Inside the Smart Cities Tech Race to Make a Durable Retrofit Electrochromic Film (EC 2.0)

Credit to Marianna Mazzucato for this excellent infographic. EC 2.0 film innovation is undergoing a similar evolution to the smart phone. What does the tech evolution of the smart phone teach us about the current race to make a durable retrofit electrochromic film (EC 2.0) for existing windows? Are the 15 underlying technologies the most important things about the iPhone? Is the integrated production vision, the “Form” (the Platonic Ideal), the real innovation? One could envision a similar infographic for the 8 known EC 2.0 projects and their core enabling technologies. Technical founders in this area incorporate learning from at least~12 specific innovation sub-fields: transparent conductor electrodes and connections, sputter chemistry, PECVD, CVD, PVD, polymer coatings, lithium ion batteries, thin films, photovoltaics, optics, spectroscopy, metrology. Most of the basic research has been funded by government.

Note: The platform opportunity for retrofitting windows (actual fenestration, not Microsoft Windows!) in the U.S. with EC 2.0 film is 19.5 billion square feet. Using a low-cost, flexible plastic substrate in a roll-to-roll process is an idea inspired by the low-cost methods used in the thin-film flexible photovoltaic industry. One analyst believes that an inexpensive EC 2.0 film can save 26 quads of energy over a 10-year period, saving 650M metric tonnes of greenhouse gas: the equivalent of 100 million barrels of oil. In Vinod Khosla’s language, it’s a “zero billion dollar market” (or, in this case, a zero five-hundred-and-thirty-five billion dollar market). EC 2.0 combines advanced materials, hardware and software innovation. It’s extremely complicated, even if the best products will be as simple and intuitive as the iPhone.

This Volume 1 research piece takes us inside the race to define the largest existing platform opportunity for the next two decades, a historical opportunity to re-define the building envelope. EC 2.0 is about “occupant-centered” design: more natural daylight, less glare, greater thermal comfort, lower energy bills. It’s about dialing the sunlight up-and-down like it’s the volume on your iPod, controlling visual environments like we do aural environments. What’s more, some of the tech may even enable Tony Stark-style data representation, where windows become screens. EC 2.0 is about adding intelligence to existing windows, like this EC 1.0 example. Here’s the vision.

What’s the big idea? Here’s a demonstration from Wisp. Disclosure: This author is a consultant to Wisp.

The focus here is on electrochromics — and specifically electrochromic retrofit film. Volumes 1 and 2 do not tackle the emerging quest to “make windows into solar walls, i.e. electricity-producing windows” (using “truly transparent solar”, quantum dots, perovskites, etc.). This is certainly another route to transforming ‘static’ windows into smart windows. Here’s a good overview of the state-of-the-state on perovskites. Our focus (EC 2.0) is on visual comfort and shading.

The complete research piece (Volumes 1 and 2) will include case studies on 8 EC 2.0 companies, an emerging startup cohort that includes ITNES,ChromoGenics, eChromic, Wisp/iGlass, NexTint, Argil, Project-X and Switch Materials. It is part of a series on light attenuation technologies for the Smart Home and EV (electric vehicle) market segments, with a specific focus on electrochromic window film (“EC 2.0”) technology. Target audiences include advanced materials and nanotech investors in Silicon Valley actively researching this investment theme for 2017 & 2018. This author welcomes dialogue with family offices, angel investment groups, micro venture capital firms ($25MM — $75MM), VC’s, corporate venture capital groups and hedge funds. This author also welcome dialogue with mentees and other graduate students at Stanford’s CEE-246 “Entrepreneurship in Civil & Environmental Engineering” course.

This piece shares Innovation Context, Innovation Valuation Ranges, and Due Diligence Questions for Investors. It’s a reference guide and companion piece, but it is not the definitive word on the subject. If you are a lead investor in a Seed, Series A or Series B investment round, there’s a 30-question Technical Due Diligence Checklist. If you are a technical founder and entrepreneur raising money, review this information to find the best partners and increase your probability of success.

Innovation Context in April 2017

Within a very short period of time — the last 8 years (2009–2017) — electrochromic researchers and product engineers (from different geographies, who mostly do not know one another and have scant familiarity with one another’s work) all proposed the same basic tech product idea: a flexible, transparent, retrofit-able electrochromic film to apply to existing windows. They wanted it to be inexpensive. Many of them knew they wanted it to be self-powered by solar (“wire-free”). They wanted it to be central to the Smart Home. It had to be durable under UV, humidity and temperature.

Some technical founders were more discerning, some better resourced, and some closer to the commercial centers of gravity — but their efforts were really similar. Small teams were creating (or re-creating) the same, or very similar, complex patterning. Metaphorically, they were all playing a fugue using the same short melody.

The design template seemed to come to each one of them with the same intensity as a religious revelation. It felt utterly unique, as though God had whispered a secret.

Is it possible that 8 different groups attempted to make the same basic product without consulting one another? Or having very minimal or superficial contact? How do we explain exceedingly similar work happening within the same rough time-frame, across different cultures, languages and geographies? Telepathy?

Is this an example of the “non-locality of mind”, à la Larry Dossey? The MRD and PRD parallels from project-to-project are stunning:

1. Low cost roll-to-roll manufacturing (enabling $30/square foot pricing)

2. Low voltage requirement, enabling zero wiring

3. Color flexibility

4. Fast switching

5. Retrofit-able to existing windows

The Founders’ respective means to get there differ somewhat. Which conductive transparent electrode materials to use? Which electrochromic composition materials to use? But the global consensus from 8 different projects on the EC 2.0 “Product Ideal” is remarkable. It’s as though there’s a Platonic Ideal of Form. The phenomena (different approaches) are shadows, mimicking the Form. They are portrayals of the Form. The important thing is the Form.

“For Plato, forms, such as beauty, are more real than any object that imitate them. Though the forms are timeless and unchanging, physical things are in a constant change of existence. Where forms are unqualified perfection, physical things are qualified and conditioned.”

Why Are We Even Talking About the iPhone?

As Marianna Mazzucato makes clear in “The Entrepreneurial State”, Steve Jobs may have had the supernatural insight to define the Whole Product and the resources to effect its existence, but he was not the inventor (which is why the U.S. Federal Government is the closest thing to an actual “inventor” of the iPhone). Because Jobs wasn’t a scientist, he didn’t have the cognitive deficits — and intellectual priorities — one observes in scientists. He didn’t fetishize technology or Intellectual Property (IP). His gifts were assemblage, packaging, editorial control. He was tech-agnostic and intellectually promiscuous. He combined vision and hustle with a personality disorder.

Was Steve Jobs original? The “Think Different” campaign is a carefully sculpted set of messages inviting comparisons between Steve Jobs and Mahatma Gandhi, Jim Henson, Maria Callas, Pablo Picasso and other 20th Century creative thinkers. But the ‘Romantic Realism’, the rational egoism we see in writing about Steve Jobs is off the mark. It’s a lazy way to think about invention.

It’s TED’s fault.

Sh*t is way more complex than the standard 15-minute infotainment format suggests. For instance, the smart phone really started with Serbians and Greeks. In terms of revelation and the iPhone, Nikola Tesla was the first one (that we know about) with a device concept combining computing and telephony. This was in 1909 and described again in 1926. Theodore Paraskevakos was the first to introduce the concepts of intelligence, data processing and visual display screens for telephones into the IP literature. That was in 1974. And then a lot of other stuff happened, and somewhere along the way General Magic became the “most important dead company in Silicon Valley” because it embodied so many of the aspects of the Platonic Ideal of Form (for a smart phone). So who is the genius? Maybe it was Marc Porat? It doesn’t matter.

It’s the wrong question.

Me Mate Rainbow Jeremy Ain’t Got No “Tech-mology”

Tech product evolution is more like language evolution. The field of Historical Linguistics includes the study of the chronological account of the birth and development of a particular word or element of a word, and includes its spread from one language to another and how it evolves in form and meaning. This is a word’s ‘etymology’. We need a new word to describe this concept of a tech product’s context and evolution. I’d like to nominate Ali G’s ridiculous neologism, “Tech-mology”, to describe this concept. Without “tech-mology”, we over-emphasize contemporary, individual contributions and we under-emphasize the collective intelligence.

Ever heard of ‘morphic resonance’? Scientists are wary of Rupert Sheldrake’s hypothesis on morphic resonance, but it’s a useful frame as we track the EC 2.0 race. What’s the idea? Basically, various perceived phenomena become more probable the more often they occur. Sounds vague and unfalsifiable, yes? Yes. So does the Vulcan Mind Meld.

This story of emerging tech innovation in the “EC 2.0” (electrochromic) advanced materials area is the story of courageous, risk-loving technical founders. Volume I provides one case study (ITNES), to demonstrate this exercise to our community. But Volume I is primarily about the trade-off between risk and price that we call valuation.

“Due Diligence is just a fancy way of saying: (1) How much of what you told me isn’t true? (2) What didn’t you tell me that I should know?”

Part II (a later April/May 2017 post, featuring Cases 2–8) will share additional profiles on 8 EC 2.0 startups/research efforts:

  1. Brian Berland, Bruce Lanning, and Michael Wayne Stowell in Littleton, Colorado (ITN Energy Systems); and
  2. Nikita Kruglikov in Novosibirsk, Siberia and Stanislav Khartov in Krasnoyarsk, Siberia and Aleksandr Timofeev in Moscow (Wisp/iGlass); and
  3. Damodar Reddy and Brian Gorgen in Santa Clara, California by way of Hyderabad, India (Argil, Inc.); and
  4. Bruce Gittleman, Dave Alie, and Nader Mahvan in Denver, Colorado (eChromic); and
  5. Greger Gregar, Johnny Engfeldt, and Niklas Lundberg in Uppsala, Sweden (Chromogenics AB); and
  6. Toby Sachs-Quintana, David Abram and Hazem Alhagrasey in Palo Alto (NexTint Film); and
  7. Mike McGehee, Dan Slotcavage, Tyler Hernandez, Michael Strand and Chris Barile in Palo Alto (EC 2.0 “X” Project); and
  8. Neil Branda, Jeremy Finden, Glen Bremner, Amir Rad, Natalie Campbell, James Senior and Patrick Chen (Switch Materials)

But before we talk about Part II, let’s talk about money and valuation.

If we believe we’re working collectively and contributing to the ‘zeitgeist’, then we’re excited to experience the flowering of ideas emerging from collective exploration — and we see idea ownership is a social construct. If we think we’re essentially alone and that the mind is contained within the brain, we give ourselves disproportionate credit.

Startup EC 2.0 Valuations and the Valley of Death

Valuation discussions are always tricky. Next to Final Status peace negotiations, they are perhaps the most severe test of the “Getting to Yes” methodology (pioneered at the Harvard Law Negotiation Project). At their worst, they devolve into ego-riddled projections of value. When this happens, they become tests of religious belief and rhetorical stridency, a kind of postfaktisch politics.

At their best, valuation discussions provide a template for further cooperation and deepening trust. They are a way for the lead investor and the entrepreneur to test how they might work together over what will likely be a 7+ year relationship.

  1. Can they separate the people (venture partner, technical founder, etc.) from the problem (valuation and terms of this round)?
  2. Can they separate their interests (growing price per share and enterprise value over 7+ years) from short-term, parochial positions (ex. showing a large uptick in an illiquid asset from 3 months earlier)?
  3. Can they invent options for mutual gain (ex. milestone-based valuation deals that reduce risk and increase valuation)?
  4. Can they insist on using objective criteria (ex. using appropriate peers to benchmark against)?
  5. Can both parties understand clearly what their options are outside of completing a particular deal, the “Best Alternative To a Negotiated Agreement” (ex. going without salaries or paying salaries late, shutting down the company, etc., for the entrepreneur; or missing out on a great financial return for the investor).

For our purposes, there are 5 groups of “comps” for valuation analysis. Skillful entrepreneurs and investors will focus on comps with defensible valuation ranges.

  1. Nano-Materials and Building Materials
  2. EC 1.0 companies
  3. EC 2.0 companies
  4. Smart Home/IoT companies
  5. Generic Seed, Series A and Series B startups

Nano-Materials and Building Materials

The former are materials (1–100 nanometers along a dimension) with unique optical, electronic or mechanical properties — for any variety of applications. The latter are new green building materials (example: a new kind of insulation). The pre-money range for “Series A” rounds in the SF Bay Area are typically $8MM — $16MM for “nano-materials”, and $3MM — $8MM for building materials. Let’s call it the “nano-premium”. It’s smaller, so it’s larger.

Example of an exception — But there are notable exceptions — and they usually happen when one large, patient Super Angel investor nurses a long R&D cycle for many, many years. In the case of Light Polymers, the company started work in 2006 but announced a $24.3MM Series A round in March 2017, 11 years later. The total paid-in capital prior to the Series A round is not clear. Here’s an interview with Marc McConnaughey, CEO of this (water-based) organic polymer company advancing the state-of-the-art for OLED displays. They are also working on an application to improve existing smart glass. As Marc says, “In smart glass applications, three layers of glass are sometimes used. We can replace the inner glass substrate with a lyotropic light management film laminated or coated on one of the other glass surfaces, replacing the function of the inner glass substrate. By replacing glass, the smart window is simpler to manufacture, lighter in handling, and lower in cost.” If we assume that this Series-A round represents 20–33% of the company’s value, the post-money valuation probable range of $72.9MM — $121.5MM looks great. But it didn’t “just happen.” They were an overnight success, after 11 years.

MIT Cleantech Report on Advanced Materials Series A Rounds — Benjamin Gaddy, Varun Sivaram and Francis O’Sullivan put together this excellent piece of research entitled “Venture Capital and Cleantech: The Wrong Model for Clean Energy Innovation.” The highlights from the 2006–2011 period under study are sobering. Half of the $25 Bln that VC’s invested was lost.

“In particular, cleantech companies developing new materials, hardware, chemicals, or processes were poorly suited for VC investment because they required significant capital, had long development timelines, were uncompetitive in commodity markets, and were unable to attract corporate acquirers. Companies developing new materials, processes, or chemicals — for solar, biofuel, battery, lighting, and other applications — returned only a sixth of the invested capital. For these companies, investors put in $764MM and $123MM was returned. Hardware integration companies, which commercialized novel ways to integrate existing hardware components, performed even more poorly, returning only 5 cents on the dollar. Investors put in $586MM and $30MM was returned.”

There are lies, damn lies, and statistics. And the IRR stats from a sample size of N=365 (Series-A rounds for “cleantech” from the 2006–2011 cycle) made every investment partnership in Silicon Valley do a lot of soul-searching. But this doesn’t mean that advanced materials and hardware innovation doesn’t happen — or that Series A or Series B investors can’t make money. It just means that they can’t naively shovel money into “hot” sectors and chillax about the outcomes. They need technical expertise, they need to be actively involved, and they need to choose companies close to revenue — and that means companies with a minimally viable product.

The Wisdoms — If you’re a technical founder aiming to minimize dilution, it’s best for you to call your company a “nano-tech” company. Even better than a “micro-tech” company. Remember, nano is smaller. Better still, call your company a “Smart Cities” play, or a “Consumer Hardware” play or “Smart Home/IoT” play (sounds like Nest!). If you’re really clever, talk about how you’re thinking about doing SaaS after a review of the Nest business model. Practice wearing khakis while saying the word “play” enough times so it doesn’t sound disingenuous. Whatever you do, don’t call your company an “advanced materials” or “cleantech” company. This is like saying that your herpes is flaring up, but that you’d really like to go on a date. (I personally love “cleantech” because, to me, it sounds like resilience. But to most investors, the term sounds like $12.5 Billion gone missing.)

If you’re an investor aiming to manage risk, it’s best to ask all of the tough questions about R&D roadmap and product readiness. Then, dispense with “taxonomy thinking” altogether like Bobby Fisher letting go of the lines on the chess board. What do you see?

“Clear the lines of men in your head.”

EC 1.0 Private Companies

These electrochromic window companies are both pre-revenue (Kinestral at a ~$300MM valuation) and post-revenue, pre-profit (~$30MM — $45MM in revenue and ~$1.2 Bln valuation). They have a different CAPEX, OPEX, TAM, SAM and SOM than the EC 2.0 firms, but their valuations are a useful marker for EC 2.0 founders. These companies benefit from having famous corporate and VC investors like Capricorn, Khosla Ventures, GE, Asahi Glass, Corning, and others. A lot of later-stage investors see value in having highly referenced investors. They pay more to be part of “branded” rounds. Let’s call it the “Schmuck Insurance-premium”.

In re: price discovery, View Glass completed a $150MM round in March 2015 at a $750MM pre-money valuation. Then they completed a $100MM raise in February 2017. The Pitchbook data:

  • March 2008: $6.6M round | $18.2M val
  • September 2009: $20.7M | $42.5M
  • February 2011: $40M | $94.6M
  • June 2012: $55M | $189.6M
  • January 2013: $100M | $502M
  • November 2013: $74.7M | $336.8M
  • August 2015: $150M | $940M
  • February 2017: $100M | $1.1B

Basically, it’s all up-and-to-the-right, minus what appears to be a hiccup in November 2013. The Series-A investors have been in the deal for 9 years. The generic profile of a company ready to go public might be: $150MM revenue run-rate, cash flow positive, with 35%+ gross margins — and a 2-year path to a $500MM run-rate. Analysts will be closely watching View Glass as a bell-weather company in the EC 1.0 sector.

A New Entrant to EC 1.0Chromogenics — This company challenges our simple taxonomy of “EC 1.0 and EC 2.0” because they make a film via a roll-to-roll EC process, but they sell this film to integrators who put it into a window and sell it to a customer. They they are making a film (“EC 2.0”), but the product is a window (“EC 1.0”). The full IPO prospectus, translated from Swedish into English, is here. Here’s their product brochure. The company is raising ~$13MM in Sweden and the stock started trading on Nasdaq First North on March 23rd, 2017. They are making rolls of foil with tint-changing properties, and shipping their electrochromic film to window manufacturers to roll out the foil between sheets of glass. The company’s technology originated from Angstrom Lab at Uppsala University –and their film can control both heat (near infrared) and light (visible light transmission). They believe they can reduce cooling costs by 50%. They have 19 patents and the 7 main shareholders includes Volvo Group. The Company has $317,000 in 2016 revenues for the first 3 quarters of 2016, consisting only of commercial demonstration projects. The Chromogenics IPO, along with the January rounds for View and Kinestral, should give analysts some feedback for their enterprise value models.

EC 2.0 Private Companies

SABIC (ex-GE Plastics division), Foxconn (largest electronics manufacturer in the world) and Flex (2nd largest electronics manufacturer in the world) are deeply interested in electrochromics. Foxconn has a head start. They are converting an existing GTOC touch panel display manufacturing facility in Taiwan to a HALIO-producing plant for Kinestral, a $100MM coup for Kinestral on the heels of several strong recent announcements. Big companies are interested in both EC 1.0 and EC 2.0 startups.

These big companies have corporate venture (CVC) programs — and some have senior scientists and engineers from electrochromic startups. In one recent conversation, a senior technical exec from a large electronics manufacturer said:

“The commercial case (for EC 2.0, the film and not the window) is a slam dunk. We get it. Don’t spend any more time talking about the commercial case. What we want to see is, ‘Will the tech work? Is it durable? What testing have you done? Will it de-laminate under temperature and humidity? Can you do the electrical connections? Can you demonstrate durable cycling in aggressive high-temperature, high-humidity environments? Is it mass-manufacturable?”

On these fronts, all 8 EC 2.0 startups need good answers. None of these companies have a commercial product yet (I’m defining “commercial product” as “a product a customer can buy and test independently right now”). This means that it’s still early in the race. The due diligence focus is in assessing tech risk.

Market risk questions tend to obscure the much more important tech risk conversation, just as the House Intelligence Committee hearings on “Who leaked?” aim to obscure the 100x more important conversation about “How exactly did Paul Manafort and Michael Flynn collaborate with the Kremlin to defeat HRC in 2016?” Startups and investors have to focus on the right question.

On the tech, these EC 2.0 startups differ by key characteristics: durability testing, Tech Readiness Level, Manufacturing Readiness Level, etc. The best way to explore this cohort is to evaluate their respective Failure Mode and Effects Analysis documents. The key questions for this cohort are: (1) What is the probability of getting to a working product within what time frame and what R&D expenditure? (2) What 3rd party data support whatever Bayesian view of #1? (3) How is tech risk appropriately priced? Valuation ranges for this cohort are from ~$3MM — $35MM pre-money, but most are in the $3MM — $10MM range. This is because of R&D timelines and follow-on capital risk. Let’s call this the “Valley of Death Deduction”.

For pre-product, pre-revenue EC 2.0 film company discussions, valuation discussions are especially tricky. But for each of these 8 companies, there are things we all know:

1. Big Markets — The markets are large and enduring. That is a vast understatement: the square footage of existing windows is 19.5 billion square feet in the U.S. alone. These are not market size-constrained ventures. The market is arguably the largest opportunity that most investors will see in their lifetimes. Investors that doubt whether the size of the market is big enough for a “venture play” aren’t serious or aren’t paying attention. For more on this issue, see “Bobos in Paradise Want ___?

2. Long R&D — The R&D development timelines are unpredictable, but more often they are long. Most of these companies work in the lab for 8–10 years before reaching a commercially testable product. The most interesting ones prove a whole lot in the lab before they emerge and raise outside money — but this requires special equipment, a large research budget, and normally a university or research affiliation. Even then, meeting the EC 1.0 requirement (ASTM-2141) for durability (a product that is durable enough to be sold commercially) can seem like scaling Mt. Whitney: 50,000 cycles at 85 Celsius and at 85% humidity. But this piece suggests that hard tech companies need to start thinking in terms of Minimally Viable Product — and stop fixating on perfection. Regardless, some companies will overcome the classic development timelines, collapsing timeframes to market — and those companies will be handsomely rewarded with favorable valuations.

3. Financing Risk — Most of the companies fail because they cannot raise further capital to fund pre-commercial R&D. For early investors, capital risk and technology risk are the biggest issues — not market risk.

Last, if investor scientists are talking to startup scientists, make sure there’s a common technical language — and that the zampolit is outside the meeting room. There’s no place for politics in science. Objective truth exists, and it’s measurable. It’s repeatable.

How do good investment teams assess tech risk for EC 2.0? The best teams hire an organic chemist or materials scientist with pre-existing expertise in EC devices — and have this person(s) do a full review of the Intellectual Property, lab-scale prototype, pilot-scale prototype, test methods, and test data, even before they study the “manufacturability”. Cyclic stability is the most important thing to study with EC devices because the degree of degradation is highly dependent on electrolyte type and operating conditions. Below are 30 starter questions (in no particular order) that the “technical due diligence guy” should ask EC research teams with prototypes.

Does the Romantic Enthusiasm Match the Science?

  1. What’s the Solar Transmittance in its bleached and colored states?
  2. What’s the Visible Transmittance in its bleached and colored states?
  3. What’s the Near-IR Reflectance in its bleached and colored states?
  4. What’s the switching speed?
  5. What’s the switching voltage?
  6. What’s the operating temperature range?
  7. Describe the optical qualities of the transparent conductor and substrate material.
  8. What’s the sheet resistance of the transparent conductor (to overcome resistive power loss and slow switching time)?
  9. Is the electrochromic material organic or inorganic?
  10. Is the structural form of the electrochromic material amorphous, crystalline, or polymeric?
  11. If you’re using partially transparent metals or metal grids for the conductive layers, what’s the reduction in transmittance?
  12. How many layers?
  13. What’s the anode and what’s the cathode material?
  14. Results of a cyclic voltammetry cycling-accelerated test? (This measures the cyclic current and voltage response of the device, generally helpful for determining the chemical changes in the electrode as it cycles)
  15. Results of square wave cycling-accelerated test? (This is used to simulate an accelerated cyclic d.c. potential response, but fixed voltage sources with controlled current are also used for testing).
  16. Results of accelerated UV exposure by use of specialized high-intensity lamps.
  17. When doing those three tests (cyclic voltammetry, square wave cycling, and UV exposure), what were the changes on spectral transmittance, reflectance and integrated optical properties)?
  18. How does humidity effect the prototype? How dependent is this device upon ambient humidity?
  19. How does oxygen effect the prototype?
  20. Is there a decline in charge capacity or optical density after long cycling?
  21. How does it fail?
  22. Film dissolution failure?
  23. Cyclic electrochemical instability?
  24. Transparent conductor etching?
  25. Gas generation? Humidity dependence?
  26. What are the secondary reactions and photo-reactions? Will its bleaching and coloration kinetics be slowed down by the effect of photochromism? Do indoor and outdoor tests show different rate kinetics?
  27. Does it show irreversible reactions with use?
  28. What colors are possible and why?
  29. Can you give us a sample so we can test it ourselves?
  30. What are the risks from materials, processes, and environments?

Confident technical teams will eagerly engage in deep technical discussions, will exhibit intense curiosity, and will look for “problem-solving” partners. They will freely share test data, show prototypes, share prototypes for independent testing, and embrace 3rd party experts to join in detailed, transparent “gate” reviews. They will not personalize critiques of their prototypes. They’ll look at investor due diligence discussions as free management consulting and thank investors dutifully for the time and effort — even if the tone is adversarial. “Yes, sir, may I have another” is an appropriate response to good criticism.

Generic Seed, Series A and Series B Startups

Both investors and entrepreneurs will use the label ‘Seed’, ‘Series A’, and ‘Series B’ rhetorically, as a kind of dog-whistle for valuation. Cap tables are typically chopped up according to certain ‘norms’ in Silicon Valley. But there are examples of EC 2.0 companies that have boot-strapped entirely through angel investment, un-priced (no max convert) convertible notes, pledge funds and other “beg, borrow and steal” strategies to achieve milestones while avoiding the class of professional VC’s. In their efforts, they’ve avoided losing near-term voting control and excessive dilution, but they’ve still paid a price on progress — and may be 12–36 months behind where they would otherwise be on commercialization. Often these startups can’t make rational long-term plans because they don’t have the balance sheet strength — and this can hurt R&D, recruiting, and pretty much everything else.

EC 2.0 companies exist within a universe of other opportunities for startup investors. They are competing for investment dollars across a range of sectors: AI/Machine Learning, Drones/Robotics, Education, Enterprise, Consumer, FinTech, VR/AR, Health/Biotech, Cannabis, and Food/Water.

The latest WSGR data show some worthwhile points for our discussion:

1. Median Series A and Seed — The Median amount raised for Series A and Seed financings in 2016 is $3.4MM, and the Median amount of post-Series A bridge loans is $1.7MM. The Median pre-money valuation for Q4 2016 Series A and Seed financings is $15.0MM (the 2nd highest quarterly figure in the past 5 years — indicating a frothy market).

2. Median Series B — The Median pre-money valuation for Q4 2016 Series B rounds is $34.3MM and the Median amount raised was $5.3MM.

3. Median Series C — The Median pre-money valuation for Q4 2016 Series C rounds is $120MM and the Median amount raised was $12.8MM.

4. Terms & Conditions (T’s & C’s) — It’s worthwhile for entrepreneurs to play with WSGR’s Term Sheet Generator and compare the average data on deal terms with the information from their Term Sheet Generator — in order to see how liquidation preferences, participating versus non-participating, dividends, anti-dilution provisions, pay-to-lay provisions, and redemption clauses, each effect outcomes.

5. Bridge to Nowhere? — Bridge loans are an attractive option for a lot of EC 2.0 film startups because they can somewhat delay the valuation discussion. The pre-Series A bridge loan median amount was $120,000 in Q4 2016 (with a median price cap of $6MM) and $1.7MM for post-Series A (with a median price cap of $25MM). 79% of bridge loan deals in 2016 had price caps — so investors taking financing risk (“Will the company be able to raise a next round?”) locked-in a valuation on 4 out of 5 deals.

Now that we’ve got money out of the way, let’s talk about innovation. The next section focuses on the first of what will be 8 EC 2.0 case studies.

But How Good Does the EC 2.0 Product Have To Be?

Here’s an editorial view just on that question. For each one of the 8 case studies in Part II, there’s an explicit or implicit Product Requirements Document. Basically, this is where companies draw the line in the sand and say they’ve succeeded in making a Minimally Viable Product (MVP). To compare each effort against one another, you’d need to know where they mark the proverbial “goal post” for product performance. For instance, one company might be OK with less-than-great optical quality (lots of haze), slower switching times, and compromised durability (loss of emissivity with UV exposure). Maybe they are focused on being first to market?

Let’s not make perfect the enemy of the good”, they say. “We’ll concentrate on the developing world.” Or, as we may remember from the Life Cereal commercial in the 1980’s, “Give it to Mikey! He’ll eat it!” And they might be correct! There are market analogies that will support a variety of commercial cases. How good was the battery in the first iPhone? Or, another example: Skype is free but sometimes the sound doesn’t work and sometimes the video stops. A dedicated video-conferencing system with perfect quality is expensive. But which dedicated video-conferencing systems do you know by name? What is good enough?

Different EC 2.0 companies have different answers. For the San Francisco Bay Area market — for both residential and commercial markets — the standards for performance (<5% emissivity loss over product lifetime, high dynamic range, no haze) and durability (10+ years) are pretty high. The great thing about EC 2.0 products is that you can tell a whole lot just by looking at the prototype. Window aesthetics is really, really hard.

CASE 1: ITNES — ITN Energy Systems

Low Cost Electrochromic Film on Plastic for Net-Zero Energy Buildings

“A market penetration of 40–50% is predicted at a price <$50/SF (incremental EC film price of $11 — $25/SF).” While this would be one of the most aggressive S-Curves (adoption /innovation diffusion curve) ever, it’s not entirely outrageous. My estimates are a bit lower, but the ITN team has a thoughtful approach too. Over the years, ITN Energy Systems developed a reputation for transitioning technologies from rigid substrates to flexible substrates that employ roll-to-roll manufacturing. ITNES’s Chief Scientist, Dr. Brian Berland, spent a lot of his career on atmospheric and surface science (atomic layer deposition for modifying porous materials for separation). He’s been at ITN since 1998 and has looked at everything from fuel cells to separation membranes, to funky ways to harness solar energy. He’s “wicked smaaht” (Casey Affleck voice).

For their EC 2.0 project (download the full project report here), ITNES benefitted from a $6MM grant from ARPA-E, arguably the preeminent funder for energy breakthroughs. The ARPA-E project ran from 2010 to 2013, and the current Director of the Office for Advanced Manufacturing at DOE’s Office of Energy Efficiency and Renewable Energy, Dr. Mark Johnson, was the Project Program Director. Interestingly, the brain trust for this project may have also included Bill Gates’s new management team for Breakthrough Energy Ventures (back when they were ARPA-E founding management): Dr. Dave Danielson and Dr. Eric Toone. Dr. Brian Berland headed up a 14-person project lasting 3 years, at twice the regular funding amount for ARPA-E projects ($6MM instead of the more customary $3MM). The project team also included the Electric Power Research Institute, Southern California Edison, Colorado School of Mines and MAG Automation and Controls. ARPA-E’s Cheryl Martin liked to say that the agency helps technologists transition from “disbelief to doubt”, indicating a focus on the earliest stages of de-risking. This project was aimed at reducing investor risk and motivating industry-led strategic partnerships for commercializing EC 2.0 film. In the end, non-government investors didn’t bite and strategic partners didn’t find enough of a de-risked technology to license, but the ITNES team still accomplished a great deal in a short period of time and showed a lot of technical leadership. They froze the project in 2014. Their objectives/deliverables included:

  1. Transitioning small-scale prototypes to 0.5 meter-wide web coater production (yield, uniformity, validated cost models, large-area prototypes, process controls)
  2. Satisfy market-driven EC performance metrics on plastic substrates (produce large-area EC film (>500 cm2) with TViz from 3–65% and high stability/durability in ASTM testing

Here’s ITNES’s ARPA-E Summary:

ITN is addressing the high cost of electrochromic windows with a new manufacturing process: roll-to-roll deposition of the film onto flexible plastic surfaces. Production of electrochromic films on plastic requires low processing temperatures and uniform film quality over large surface areas. ITN is overcoming these challenges using its previous experience in growing flexible thin-film solar cells and batteries. By developing sensor-based controls, ITN’s roll-to-roll manufacturing process yields more film over a larger area than traditional film deposition methods. Evaluating deposition processes from a control standpoint ultimately strengthens the ability for ITN to handle unanticipated deviations quickly and efficiently, enabling more consistent large-volume production. The team is currently moving from small-scale prototypes into pilot-scale production to validate roll-to-roll manufacturability and produce scaled prototypes that can be proven in simulated operating conditions. Electrochromic plastic films could also open new markets in building retrofit applications, vastly expanding the potential energy savings.

How far did they get in their development? In short, pretty far. They went from a TRL 4 to a TRL 6, and from a MRL of 3–4 to a MRL of 6–7. Not commercial, but well beyond Cheryl Martin’s “from disbelief to doubt” metric.

The word clouds of EC 2.0 emphasize process engineering concepts. This word cloud is based on ITNES’s ARPA-E report.


The team demonstrated that they could successfully deposit material (nickel oxide, tungsten oxide) and demonstrated their EC device on PET film. They had 4 deposition zones, one for each layer: transparent conductor, electrolyte, ion storage, and electrochromic layer). They achieved a dynamic range of between 5% — 65% visible transmission in solid state devices — and they could modulate TViz from 3–70% with composite devices. They could use PIB material to laminate the film to glass. (PIB is a polyisobutylene-based sealant that exhibits excellent long-term stability and remains permanently flexible, even at low temperatures.) They could do the roll-to-roll coating, laser scribing to pattern them after-the-fact, and the PVB lamination. Their leakage current was very low (at the time Sage wasn’t trying to make a device leak-proof): less than 100 mW/ft2 for devices >600cm2. They produced ITO transparent conductive layers with high conductivity (~25/Ohms/square) and transmission (~87% TViz) at 1/3rd the thickness of poorer performing commercially purchased ITO employed for initial development.

For manufacturing, they demonstrated roll-to-roll production of solid state EC devices on PET substrates: up to 2 meters of continuous production of individual layers on 0.5 meters wide web. They dramatically increased the deposition rate and decreased the thickness of the electrolyte layer (lowering manufacturing per unit cost). They had some challenges achieving uniformity of all the layers over large areas.

The ITNES team emphasizes intelligent process controls, which is a kind of institutional knowledge you only achieve by running pilot production. In the EC 2.0 world, when you change one parameter, you often experience unintended consequences. One thing can change a bunch of other things. The ITNES team wrestled with this on a lot of fronts, and probably developed a lot of wisdom. They appear to understand a lot of reflexive relationships that aren’t obvious.


Most impressive, the team demonstrated “tens of thousands” of cycles in high temperature, high humidity environments. They were using both a thin-film protective barrier coating and then a commercial barrier film (for 40 C, 100% relative humidity environments). Their suspended film architectures withstood >20,000 cycles and their product (laminated to glass) on the outside of the IGU >10,000 cycles.


Pretty good dynamic range (3% — 65%), pretty low switch time (2 minutes, versus 22 minutes for EC 1.0), and awesome durability versus ASTM 2142–02 protocols that include heat and solar exposure. This looks like a product that can be warrantied for 10 years. 20,000 cycles with essentially no switching range is nothing to sneeze at!

Their energy performance diagnostic was stymied by challenges in developing devices that use their Gen 2 ion storage layers. ITNES couldn’t measure reflectance, transmission, etc., and these properties are required to be able to calculate Solar Heat Gain Co-efficient, U-Value, etc.

Challenges & Wrap Thoughts

They solved a lot in the project, but there were still some further technical challenges — even after $6MM in productive work. Could the plastic be sufficiently cleaned so there aren’t electric shorts? Could they make a much better deposition tool? Could they increase the switching times (this has to do with the transparent conductors and the distance between the but bars and ITO; they started at 50–60 Ohm per conductivity and arrived at 15–17 at the end of the project). They were able to move from a relatively low active area (for the transparent conductive electrode) to an active area at least 1/2 meter wide, making it big enough for a residential application.

Everyone is interested in cost. The $33/square foot number is pretty good. It’s not clear whether they are using a pricier flexible moisture barrier to demonstrate suitability for retrofit apps. What’s the Water Vapor Transmission Rate (WVTR)? Do you need a moisture barrier that’s just food packaging quality? Or a much more expensive, much more robust ultra-barrier film, like the kind you need for Organic Light-Emitting Diodes?

Of course, timing is everything and the environment in September 2013 was very different than the environment in April 2017. On the market challenge front, both the Department of Energy and the commercial players were not thrilled with anything ‘electrochromic’ at the time. The window market was in dire straights from 2009–2012. The team had shown what the material sets were able to achieve (with tweaking of band gap for color), but the larger companies they tried to market to wanted to see a more mature product. ITNES had a nice market report from EPRI on how to work with utilities.

In sum, this team did a lot of things right and they showed some real technical chops. Big thanks to ARPA-E for making their taxpayer-funded work available to other innovators in the field.


Volume II explores the efforts of several EC 2.0 companies and their quests to make a flexible, durable electrochromic film for retrofitting existing windows. Our target efforts are below and if you have specific thoughts on any of them, feel free to e-mail me at luke at pustejovsky dot ventures.

CASE 2: E-Chromic

CASE 3: Argil, Inc.

CASE 4: Chromogenics

CASE 5: Wisp/iGlass

CASE 6: Project-X

CASE 7: NexTint

Closing Note — All Our Patent Are Belong To You

Some will inevitably resist comparisons — and even the whole notion of making comparisons. They’ll want to reinforce their uniqueness, and certify their private conversation with God when they were given secret instructions about how the universe works. But if you believe — on the metaphysical level — that we’re working collectively, and that our work is contributing to the zeitgeist, then you want to see the flowering of ideas emerging from this collective exploration.

If this is you, then you’re on the “open” part of the Open Innovation Spectrum. You are the kind of person who would decide to open up your IP to all electric vehicle manufacturers, as Tesla did. For you, you’re very fortunate. Because you know that the story itself is bigger than any of its individual protagonists. The evolution is the point.

Luke Pustejovsky

Written by

Luke is an entrepreneur in the San Francisco Bay Area, focused on energy, advanced materials, climate change and “hard tech” related ventures.