Elemental Machines, a venture based in Boston and San Francisco, has come out of stealth mode. The startup says it’s raised $2.5 million in seed from investors including Founders Fund’s FF Angel, PayPal co-founder Max Levchin and Project 11 Ventures. And now it’s ready to change the way our world does science, providing the infrastructure that will ensure experiment reproducibility for researchers.
The founding team includes Sridhar Iyengar and Sonny Vu, previously the founder of wearables maker Misfit and medical device firm AgaMatrix. Iyengar (who has a PhD in biochemistry) and Vu are joined by co-founders Elicia Wong and Gary Tsai, and they’re planning on bringing their sensors (called “elements”) and cloud-based software to scientists. The goal is for scientists to gain intelligence into any variability in the physical world that may be impacting their experiments, research or reproducibility. Ultimately they want to provide intelligent systems as part of Science 2.0 — giving scientists and researchers the ability understand the contextual variables that are affecting their work and improve reproducibility.
“Why does it take an entire decade to bring a new medicine to market?”
Iyengar explained, “We’re looking at big problems to solve, as any startup does, and we stumbled on solving one: the process of problem-solving itself.”
In the past, Iyengar had encountered his own issues as he was trying to problem-solve in the lab.
He went on to elaborate:
Why does it take an entire decade to bring a new medicine to market? It shouldn’t take that long, but the concept of taking a long time keeps popping up over and over again. In life sciences, when you’re playing with atoms and molecules, it takes a long time to do experiments, so why not help scientists invent and discover things faster?
When you do experiments, silly things happen and it throws you off. From my personal experience, one time I spent 6 months figuring out what was wrong with a formulation. Turns out, the humidity on the day I made the formulation threw it off. It took 6 months for me to figure that out. Guess how long it took me to fix that formulation once I knew what was wrong? Less than a minute.
Usually, the fix in these situations is simple. But the process of pinning down what made an experiment go awry eats up scientists’ time. With the startup’s EM Suite, sensors will measure the most common sources of experimental variability: temperature, humidity, vibration, light and instrumentation. The raw data will be used to extract metadata, perform analytics and present them to scientists in a comprehensive way so they can determine whether these factors interfered with their experimental results.
To put Elemental Machines into context, Iyengar gave an analogy:
Elemental Machines helps scientists debug the physical world…Imagine trying to write software on a program that changes every day. That’s biology. Every time you do an experiment, you’re following a protocol, but on a different system because of the physical world…But if software crashes, a developer doesn’t go back and write all of the code from scratch. They go back, see where it went wrong, debug it and fix it. Scientific protocol is basically pseudo-code and the same thing can be done.
Essentially, the startup will be saving scientists months of guess and check whenever they’re figuring out what went wrong with an experiment. And as Elemental Machines emerges from stealth, it’s looking to expand its team and gain wider acceptance throughout the scientific community.
Image via Elemental Machines.