Startup of the Week: Citrine Informatics

The Innovator

By Chris O’Brien

Citrine Informatics, a World Economic Forum Technology Pioneer, is using big data and AI to make creating new materials more efficient. The Silicon Valley company believes such an approach can bridge a gap between traditional, but slow, approaches to material science and the hunger of consumer electronics companies to deliver breakthrough products every couple of years.

“These product companies were starting to push the limits in terms of what they were making their products out of,” says Citrine co-founder and CEO Greg Mulholland. “And that led to a crunch in materials industry. It used to take ten to twenty years for new materials to be developed and these product companies wanted it in 36 months.”

This push to inject the power of software and silicon into the traditional manufacturing industry is variously known as the “4th Industrial Revolution” or “Manufacturing 4.0.” The trend includes technologies such as 3D printing for industry, the Internet of Things, robotics, material science, and AI. The Future of Production was a major theme at the recent World Economic Forum in Davos, Switzerland.

Last June, Citrine raised $7.6 million in venture capital, its first round, from a series of high-profile investors, including Data CollectiveInnovation Endeavors, and Prelude Ventures, LLC. It now counts among its partners and clients 3M, Panasonic, and AutoDesk.

Founded in 2013, Citrine was the result of a collaboration between founders with backgrounds in material sciences and others who came from the AI world.

“We realized that there there is a big change that’s coming in the material science industry,” Mulholland said. “One that’s pretty fundamental.”

Continue reading…

Panasonic Set to Optimize Development With Materials Informatics

Panasonic Corp. (TYO:6752) aims to increase the efficiency of its new materials development efforts by using the materials informatics (MI) technology of Citrine Informatics Inc., a Silicon Valley venture firm.

By optimizing its use of enormous volumes of accumulated proprietary data, Panasonic expects to be able to halve development time and complement its scarce development resources.

Part of a company-wide strategy aimed at future growth, the MI digitization project will also deepen Panasonic’s involvement in the concentration of world-class technologies in the Silicon Valley area.

Continue reading…

The AI Company That Helps Boeing Cook New Metals for Jets

Wired Magazine

By Sophia Chen

At HRL Laboratories in Malibu, California, materials scientist Hunter Martin and his team load a grey powder as fine as confectioner’s sugar into a machine. They’ve curated the powder recipe—mostly aluminum, blended with some other elements—down to the atom. The machine, a 3-D metal printer, lays the powder down a single dusting at time, while a laser overhead welds the layers together. Over several hours, the machine prints a small block the size of brownie.

HRL’s parent companies, Boeing and General Motors, want to 3-D print intricate metal parts in mass for their sleek new generation of cars and planes. Airbus has already installed the first-ever 3-D printed metal part on a commercial airplane, a bracket that attaches to its wings. But the tech is limited by the quality of today’s metal powders, says Martin. Most useful alloys aren’t printable because the atoms in the powder grains don’t stack correctly—leading to a weak, brittle weld.

So Martin’s group, which largely works at Boeing and GM’s forward-thinking HRL’s Sensors and Materials Laboratory, figured out how to alter the recipe of a strong alloy so it was compatible with a 3-D printer. Their secret weapon: a machine learning software made by Bay Area-based company, Citrine Informatics. It turns out, algorithms can learn enough chemistry to figure out what materials Boeing should use in their next airplane body.

Martin’s test block took more than 2 years of work. Scanning through the periodic table, his team came up with 10 million possible recipes for improving the powder. Then, they had to figure out which ones to try to make—using Citrine’s machine learning algorithms.

When companies upgrade their products—the next Prius, smartphone, or raincoat—they first consider how to upgrade the materials they’re made of. They could be improving quality, like making a harder glass for the iPhone, or figuring out how to make a cheaper battery. “Everything has to start with, what are we going to make it out of?” says materials scientist Liz Holm of Carnegie Mellon University, who has collaborated with Citrine in the past.

 But historically, this process takes forever. If you were trying to make a more efficient LED, you’d use your years of materials science experience to pick an initial semiconductor recipe, and then you’d tweak it ad nauseum for years, until the material fit all your criteria. “You know the scientific method,” says Greg Mulholland, the CEO of Citrine. “You come up with a hypothesis; you test it; you conclude something. And you start over.”

So in 2013, when Mulholland was still in business school, he and Citrine co-founders Bryce Meredig and Kyle Michel thought they could speed up that process. A crucial step is to pick the first recipe in the right ballpark, which usually takes the touch of an experienced researcher who has worked with similar materials for years. But instead of relying on one scientist’s limited experience, why not ask an algorithm fed with decades of experimental data?

To create these algorithms, they had to trawl for the data from those decades of experiments. They wrote software to scan and convert the data printed in heavy reference books from another era. They fed their algorithms the results of supercomputer simulations of exotic crystals. They built a friendly user interface, where a researcher can select from drop-down menus and toggle buttons to describe the type of material they want. Other than HRL, the Citrine team has partnered with clients such as Panasonic, Darpa, and various national labs in the last four years.

But even still, materials science projects suffer from a lack of data. “We have to do some creative things to really make the most of the data available,” says Mulholland. Unlike, say, the algorithms underpinning Google Translate, which are trained with millions of words, you might only have a thousand data points or fewer for a class of materials. Some companies want to work with materials only discovered a few years ago. To give the algorithms more to work with, Mulholland’s team teaches the algorithms general rules about physics and chemistry.

Sometimes they even have to resort to handwritten data. “There are times when we have to scan papers and notebooks from our customers, which is truly awful,” says Mulholland. “The norm is close to what my lab notebooks used to look like. It’s a series of hard-to-read notes, interspersed with chemicals dripped onto pages.”

 Luckily, they didn’t have to go that far with Martin’s group. Martin found out about Citrine when Meredig, Citrine’s chief science officer, gave a talk at his graduate school. They figured out that Citrine could predict what atoms to add to their alloy to improve weldability. For example, the algorithm could ballpark the optimal size of the atoms and and the type of chemical bonds they’d need to form. The software helped Martin’s team rule out most of the 10 million proposed recipes to a manageable 100. Conventionally, this process would have taken place in the lab over iterations of experiments. “What would’ve taken years, it narrowed it down to days,” Martin says.Using those new powder formulations, they printed several prototype blocks and tested their strength. When they examined the blocks under microscopes and pulled them with thousands of pounds of force, they passed the test.

But as smart as the Citrine software is, it’s not going to replace human expertise, says William Paul King of the University of Illinois at Urbana-Champaign, who was not involved in the research. Martin’s team couldn’t just tell the software, “Fix this unweldable powder!” They had to tell the algorithm explicitly what chemical properties they were looking for. “It required significant expertise from them,” says King.

Instead, it makes it possible for materials scientists to use more of the institutional knowledge they’ve built for decades. “It shouldn’t take 100 years to have really advanced answers to a lot of these materials science questions,” says Mulholland. “It should take five to 10 years. Or shorter than that in some cases.” In answering Martin’s 3-D printing question—Citrine knocked that down to days.

Learn more here


Citrine Informatics Hires Senior Global Sales Leaders

REDWOOD CITY, Calif.–(BUSINESS WIRE)–Citrine Informatics, the chemicals and materials artificial intelligence (AI) platform, today announced the expansion of its sales organization through the appointment of two seasoned sales leaders with years of experience growing worldwide technology field operations and servicing the needs of Fortune 1000 and manufacturing customers. Deepak Ghodke and Brian Gillespie have both been appointed as Vice President of Sales.

Continue reading…

Citrination Plays Key Role in Creating 3d-Printable Aluminum Alloys

With the help of Citrine Informatics’ Citrination platform, scientists at University of California Santa Barbara and HRL Laboratories, a joint venture of General Motors and Boeing, have solved a long-standing challenge hindering the wide use of aluminum alloys in 3d printing. In particular, the high-strength aluminum alloys of relevance to the aerospace and automotive industries have notoriously remained impossible to 3d print due to a stubborn problem called hot cracking. In a study published in Nature, one of the world’s most prestigious scientific journals, the researchers employ a combination of metallurgical expertise and Citrination to design a robust solution to hot cracking.

Describing the breakthrough, John Martin, a scientist at HRL Laboratories, says, “We took an unweldable old metal and made it weldable.” How? The team started with a strong scientific hypothesis grounded in well-established metallurgical principles. They reasoned that the introduction of highly tailored grain refiners, a tried-and-true route to improve processability and properties of alloys, could make existing commercial aluminum alloys amenable to 3d printing.

However, testing this hypothesis presented an enormous challenge. The team would have to sift through millions of combinations of candidate grain refiners and target aluminum alloys to find ideal matches. This crucial step, for which no other software solution existed, is where Citrination came in. Brennan Yahata, another scientist on the study, notes, “Using informatics was key. The point of using [Citrination] was to do a selective approach to … find the materials with the exact properties we needed. Once we told them what to look for, their big data analysis narrowed the field of available materials from hundreds of thousands to a select few. We went from a haystack to a handful of possible needles.” Specifically, the authors of the study used Citrination to search millions of candidate materials combinations to find optimal sets of four key properties for which they were screening. When they tested these optimal combinations in the laboratory, the scientists found that the resulting parts had dramatically higher strength than had been demonstrated previously in 3d printed aluminum.

Martin calls this work a “true optimization from the atomic scale up to the component scale,” which has long been a dream of materials design. The study opens new frontiers in the use of aluminum alloys in 3d printing, and demonstrates the great potential of informatics in helping to unlock previously-intractable materials design problems.

  • B Meredig, Chief Science Officer 


Further reading:

·       Engineers 3-D print high-strength aluminum, solve ages-old welding problem using nanoparticles []

·       JH Martin et al., 3D printing of high-strength aluminium alloys, Nature (2017)

Engineers 3-D print high-strength aluminum, solve ages-old welding problem using nanoparticles

HRL Laboratories has made a breakthrough in metallurgy with the announcement that researchers at the famous facility have developed a technique for successfully 3D printing high-strength aluminum alloys—including types Al7075 and Al6061—that opens the door to additive manufacturing of engineering-relevant alloys. These alloys are very desirable for aircraft and automobile parts and have been among thousands that were not amenable to additive manufacturing—3D printing—a difficulty that has been solved by the HRL researchers. An added benefit is that their method can be applied to additional alloy families such as high-strength steels and nickel-based superalloys difficult to process currently in additive manufacturing.

Continue reading…

3D Printing of High-Strength Aluminium Alloys

John H. MartinBrennan D. YahataJacob M. Hundley, Justin A. MayerTobias A. Schaedler,  Tresa M. Pollock

Metal-based additive manufacturing, or three-dimensional (3D) printing, is a potentially disruptive technology across multiple industries, including the aerospace, biomedical and automotive industries. Building up metal components layer by layer increases design freedom and manufacturing flexibility, thereby enabling complex geometries, increased product customization and shorter time to market, while eliminating traditional economy-of-scale constraints. However, currently only a few alloys, the most relevant being AlSi10Mg, TiAl6V4, CoCr and Inconel 718, can be reliably printed12; the vast majority of the more than 5,500 alloys in use today cannot be additively manufactured because the melting and solidification dynamics during the printing process lead to intolerable microstructures with large columnar grains and periodic cracks345. Here we demonstrate that these issues can be resolved by introducing nanoparticles of nucleants that control solidification during additive manufacturing.

Continue reading…