At first blush, a company like Tesla Motors might not appear to have much in common with a company like Dow. But to materials scientists like Bryce Meredig and Greg Mulholland, both MBA ’14, those two companies — and more like them — are simply in the business of inventing and producing materials that the world has never seen before.
In 2013, Meredig and Mulholland co-founded Citrine Informatics, which now has five employees and plans to expand. Citrine’s software combines big-data analysis with advanced computer learning to help those companies with every phase of their materials development, from initial R&D through product design to manufacturing at scale. It’s a long, fraught process, and often a company doesn’t know even where to start. “A lot of companies don’t know what materials they need,” Mulholland says. “They just know they need to solve a problem.”
That space of uncertainty presents an opportunity for Citrine, which has raised an undisclosed amount of funding from investors, including Innovation Endeavors and XSeed Capital. “Materials problems are everywhere, and they’re some of the hardest problems that humanity faces,” says Mulholland. “We think we can make them much easier.”
The amount of advanced materials work that goes into everything from the smartphones in our pockets to the aircraft in the sky is tremendous. Mulholland witnessed all the pain points of materials development as an engineer and then director of operations at Kyma Technologies, a company that specializes in growing and fabricating advanced semiconductor materials. From trying to understand what a material is doing on a fundamental level through having it perform as expected to producing it at scale, “it turns out that making things is really hard,” he says.
Citing a study by the National Academy of Sciences, he estimates that, due to the onerous cycles of testing and failing and tweaking and retesting, the process of developing a material and being able to insert it into a product takes about 20 years. “It’s one of the slowest product pipelines on Earth,” he says.
Citrine’s approach is to use data to dramatically accelerate the entire pipeline. There are huge amounts of data available about how various materials perform under different circumstances, but they all live in different places and in different guises.
Citrine’s algorithms mine and aggregate all the available data — from academic literature to public databases to a company’s own proprietary research and documentation. The models can then identify a list of candidate materials with the necessary properties for a product — as well as predict the most likely ways that those materials might fail, giving the company a chance to ameliorate problems during the R&D process.
“What we do represents a big change in the status quo,” Meredig says. “A lot of the scientists in these industries don’t have a natural inclination to turn to software to solve these problems.”
Citrine first put its technology to work in a pilot program with researchers at UC Santa Barbara to convert waste heat into electricity. They ingested data about a wide variety of known thermoelectric materials and applied analytic engines to identify new candidate materials that were well outside of the list of usual suspects. The researchers made one of those materials and found that it worked. “They said never in a million years would they have thought about it on their own,” Mulholland says.
While Citrine’s early success has come in the discovery side, mostly dealing with thermoelectrics and solid-state lighting, the company is now focusing more on so-called smart manufacturing. “That’s a huge opportunity,” Meredig says. “Helping companies use their materials data to optimize manufacturing is the direction we’re headed in now.”
In fact, Mulholland says, the future of manufacturing lies in promising technologies like 3D printing. Look forward 50 years, and 3D printing will be a major component of everything from the products we use every day to advances in aerospace technology. And yet, one of the major challenges in taking 3D printing to the next level will be in understanding how those printed materials will perform in extreme conditions. A data-driven approach like Citrine’s could help clear those hurdles.
Citrine Informatics received an early seed grant from Stanford’s TomKat Center for Sustainable Energy, which provides grants and mentorship to develop renewable energy technologies and policies. While the potential applications for Citrine’s platform are broad — Mulholland estimates that at least a third of the companies on the S&P 500 either make or rely on advanced materials — the work is particularly relevant in advancing the mission of the TomKat Center.
“The biggest energy issues facing us today are materials issues,” says Mulholland.
Look at the batteries being touted by Tesla, he says. They are a pure materials problem, literally three materials layered on top of each other in an interesting way. Solar energy has a significant materials component, from the composition of the panels themselves to making the manufacturing process more efficient for U.S. firms to better compete with market-leader China. Even considering the realm of fossil fuels, combustion is more efficient the hotter that fuels burn, so high-temperature alloys can have a significant impact in reducing the amount of carbon emissions. Even problems like California’s water shortage can be tackled by novel materials, Meredig says, as advanced membranes could unlock innovative technologies for seawater desalination.
Energy consumption, water usage, and the next generation of devices are all being led by the materials that enable them, say Meredig and Mulholland. And while there is often a great deal of headline-grabbing when a promising new material pops up — just look at graphene — the team at Citrine is focused on enabling the work involved in going from a cool idea to a scalable product.
“Materials development can be such a boon for society, but there is also a tendency to chase bright, shiny objects,” Meredig says. “What we care about as a company is putting materials into a car, putting new solar materials on people’s roofs. Until that happens, we haven’t succeeded.”
This article was originally published in Insights by Stanford Business