Kevin Coakley sold his first solar company, ThinSilicon, to China Solar in 2009, in a deal he described as a “catching the falling knife-type acquisition.”
The company was quickly running out of money, Coakley said, and China Solar was offering enough for the super-thin wafer business to allow him to move on. His next venture? A flexible circuit for solar panels — or a cell link, if you will — that created more open area for light to penetrate a cell by moving wiring to the back.
Coakley remembers that as a weird time in the clean energy world. Solar was “pretty much radioactive at that point,” he said. “There wasn’t a whole lot of interest from the investment company in funding my idea.”
His design offered marginal efficiency gains to solar panels, but upped the cost of manufacturing. And Silicon Valley simply wasn’t interested. But in 2013, the Department of Energy’s Solar Energy Technologies Office agreed to give the new company, dubbed CelLink, a $750,000 grant to iron out the circuit’s design and cut test patterns.
At the time, said SETO deputy director Garrett Nilsen, there was still a lot of “white space” in the solar industry — nobody quite knew which tech would rise to the top, or when. So SETO was essentially funding anything and everything that might be a key solution for the market when it reached maturity.
“Finding ways to squeeze out more efficiency was a really big deal in the industry at that time,” he said, and that’s exactly what CelLink was proposing. And it was doing it at a lower price point than Sunpower, which in 2013 was dominating manufacturing of wire harnesses.
“The stars seemed to align, the price point seemed to work,” Nilsen added.
But before CelLink had commercialized its flexible circuit tech, the cost of solar fell off a cliff. The company never sold a single back-contact harness to the solar industry.
Solar’s price plunge
Making CelLink work from a financial standpoint was always going to require a “very aggressive” cost model, Coakley said.
They weren’t oblivious to the fact that the cost of solar was coming down; between the time when he sold ThinSilicon and when SETO agreed to back CelLink, utility-scale PV dropped from $4.75 to just over $2 per watt.
But CelLink “miscalculated that trajectory,” Coakley said. By 2015, the price was well below $2 per watt, and entire panels were rapidly becoming cheaper than CelLink’s single component.
“It came down so fast that we really knew implicitly that we were going to have to pivot somehow,” Coakley said. “It was kind of a soul crusher of a feeling in the solar industry — this is coming down so fast that there’s not a real chance to break through.”
A forced pivot is generally not a good sign for a startup.
“It implies stress,” explained James Cross, co-head of private investing at Franklin Templeton, which joined CelLink’s Series B investment in 2019. “It means they may have to take a down round or a bridge or a convert, maybe do layoffs, or they’ve got to retool the company.”
But CelLink was a unique case, Cross said. When the company decided to leave solar behind and instead rework its equipment for the electric vehicle market in 2015, investors saw the pivot as “an elevation,” he said.
And, thanks to what Coakley described as essentially “good fortune,” CelLink fast found customers. At a 2015 meeting, a “well-known EV maker in the Bay Area” (yes, that one) suggested using the flex harness to replace traditional wire harnessing in EVs.
The EV pivot
That was the start of CelLink’s new phase.
“It was just a much better fit for our product,” Coakley said, attributing much of that fit to the customized nature of vehicle design, compared to the solar industry where things were largely becoming commoditized by 2015. “When you have customization, that means you can have innovation.”
The size and weight savings CelLink’s harness offered were very attractive to the company’s automaker patron.
On the design side, saving space is a big deal, Coakley said, because people want a roomy cabin to sit in, and “big bulky wire harnesses take up a ton of space.” CelLink’s flex harness is flat, and saves precious millimeters of space, he added.
And on the efficiency side, weight greatly impacts driving range — the lighter the car, the further it can go on a single charge. CelLink’s flex harness is around 70% lighter than traditional wiring, in part because a flat harness dissipates heat better than a round wire does, requiring less wire overall, which Coakley said saves “kilograms” in weight.
When you have customization, that means you can have innovation.
Those are sentiments echoed by Cross, at Franklin Templeton, who said they agreed to back CelLink only after spending a lot of time with auto industry OEMs to determine whether the tech was a “nice to have” or had the potential to be transformative. He described the performance improvement as “very significant.”
It took five years to get CelLink’s harness on the road with their first EV partner, Coakley said. But if that seems like a long time, he caveated, “you should see how slowly some of these other automakers move.”
Nixing the solar blinders
That five year process of building a foundation in the auto industry brought CelLink back to DOE for funding; this time to the Loan Programs Office.
In late April, LPO finalized a $362 million loan to help CelLink build a manufacturing facility in Georgetown, Texas, where the eventual plan is to churn out enough flex harnesses for 2.7 million EVs every year.
That factory, which DOE hopes will create more than 1,200 permanent jobs, is the culmination of the type of innovation trajectory the agency is designed to support, said LPO director Jigar Shah.
And furthermore, it’s an indicator that the U.S. is entering the next era of clean tech.
In “cleantech 1.0,” manufacturing like this essentially had to be done in China, Shah said. A few years ago, “no one in a million years” thought the kind of equipment manufacturing that CelLink is planning for could be done in the U.S.
To be clear, it’s not unusual for a startup receiving DOE funding to realize that their initial target market isn’t the best for their tech, Shah said.
“People pivot all the time,” he added. “It’s more that, once they pivot, they don’t really think they can manufacture here.”
That’s what makes CelLink’s arc unusual — and in the eyes of LPO, ideal. Rather than licensing its technology for another company to produce abroad, CelLink is doubling down on making it domestically.
And even though CelLink has left the solar sphere, Nilsen said that’s just fine with him. SETO routinely encounters companies and technologies that are at such early stages, they “know the final application might end up somewhere else.”
SETO was looking for solar solutions, Nilsen added, but it shouldn’t ever be “solar or bust.”
