You’ve heard of the smart home. You probably haven’t heard of the smart lab. Until now, that is.
The idea is analogous to a connected home: you hook up a science lab with sensors and software to track things like temperature, humidity, and which piece of equipment was used when. Why? If you’ve ever spent three months debugging what went wrong with a chemistry experiment, you know why.
Sridhar Iyengar was that guy in a previous life. It turned out (sparing the details) that his chemical formulation was unstable because of the humidity in the room on the day of the experiment. Having that kind of experience in science may have helped motivate him to become an entrepreneur.
Iyengar (pictured above) is best known for co-founding Misfit Wearables, a fitness-tracking startup that was acquired by Fossil Group last fall, and AgaMatrix, a medical device and mobile-app company focused on glucose monitoring for diabetes.
Now Iyengar and his Misfit co-founder Sonny Vu—together with Elicia Wong and Gary Tsai—have started a new company based in Cambridge, MA, called Elemental Machines. For Iyengar, the startup’s CEO, things have come full circle, back to that lab bench experience.
Elemental Machines says today it has raised $2.5 million in seed funding from investors including FF Angel (run by Founders Fund), Max Levchin, Project 11 Ventures, 2M Companies, and Rock Health. Most of the money was raised last year, and the round closed in August. (Founders Fund and Levchin were also investors in Misfit.)
“Elemental Machines is at the intersection of life science and machine learning—two of the most exciting fields to be in right now,” says Levchin, the co-founder of PayPal and CEO of Affirm.
It’s safe to say the investors are betting on the founders’ track record and their vision. Which is, as Iyengar puts it: “How do we help people invent things faster?”
More specifically, Elemental is providing a system of software and sensors to monitor and analyze things like instrument usage and lab conditions—heat, humidity, lighting, vibrations, gases—with the goal of helping scientists debug experimental problems and get repeatable results more quickly.
The company’s initial market is life sciences research and development, from startups to universities to big pharma. Think drug discovery, where the time from concept to marketing a new drug can be 10 years, and costs can be astronomical.
In the past decade, a lot of people have talked up automated approaches to running scientific experiments, and new methodologies to speed up drug research and development. What’s interesting about Elemental is that it’s going after simple but prevalent aspects of a bigger problem in science.
And that is that crucial experiments are often delayed by things that seem trivial in retrospect. “I talked to my friends who worked in labs,” Iyengar says. “Everyone had a story to tell.” One scientist’s polymer was unstable because of ultraviolet light coming through a nearby window, he says; that took six months to debug. Another friend who worked at a pharmaceutical company was testing drug candidates in mice. The results were one failure after another, for months, until someone figured out that the lab next door was being renovated, and after-hours construction was keeping the mice awake and stressing them out.
Not that sleepless mice are the reason for slow drug pipelines, but you get the idea. Iyengar says if Elemental’s system can help shave a year off the drug development process for a given company, say, “what’s really compelling is the economics.”
Elemental’s approach boils down to applying methods from consumer-health IT (distributed sensors and apps) and manufacturing (process control) to early-stage research and development, Iyengar says. The company’s software and dashboard interface (see below) help scientists detect anomalies in experiments and “pull out hidden trends and patterns” in workflows, he adds, sort of like “a debugger for the physical world.”
Iyengar points out that labs doing cutting-edge science are often surprisingly low-tech when it comes to keeping records and monitoring experiments—think pen and paper, three-ring binders, and (maybe) remote login systems. That’s changing, though it won’t happen overnight. “People are going to get wired up, but we’re one of the first,” he says, “to be doing it in an experimental context.”
It’s not a new idea, in other words. But Elemental seems like a shrewd play to get in on the ground floor of a movement to “connect” everything from homes to cars to businesses, and make them smarter. Its grand ambition may be to disrupt science, but it will certainly settle for a big return on investment from biotech companies and academic groups. That its product and market require expertise in software, hardware, and life sciences may be a key differentiator for the startup.
What lessons does Iyengar bring from his previous companies? From Misfit, he says, it’s the importance of a “good, easy, simple” user interface and user experience for a given product. “People want to plug in and use it,” he says. From AgaMatrix, it’s that “you don’t need a million sensors” to make a difference in life sciences.
For a mid-size customer’s lab, he says, it could take 50 to 100 sensors to make a big impact. And Elemental stresses that its main value lies in the software modeling, machine learning, and analytics that try to make sense of the measurements and help customers run more effective experiments. In that regard, of course, the company will need to deliver.
One prominent biotech CEO, who didn’t want to be quoted by name because he’s not familiar with Elemental, says “the idea is interesting,” and that “it is all in the details to see if it really works and drives value.”
Elemental Machines has 15 employees split between Cambridge and Burlingame, CA. “We are actively recruiting and hiring,” Iyengar says. He adds that the company’s first product is now in beta testing. Kendall Square’s LabCentral startup facility is an early customer (others can’t be named yet). Elemental says its product will launch publicly later this year.
