Here’s a strange truth about modern science. In fact, AI can now dream up experiments faster than any lab can actually run them. So the slow part isn’t the thinking anymore. It’s the pipetting. That’s the gap autonomous lab robots are trying to close, and this week one of the best-known names in the field raised fresh money to do it.
The company is Opentrons, and it just announced a $40 million funding round. So let’s look at what that buys, why it matters and why you might care.
So what actually happened?
Opentrons Labworks raised a $40 million Series E. That’s a late-stage round, the fifth lettered one after Series A through D. Healthcare investor Avego led it, and BroadOak Capital Partners also put money in, according to GEN.
First, the company says the cash will go into robotics, software and data tools. It also wants AI that runs experiments directly, faster protocol development and a push into new markets and regions. A protocol, by the way, is just the step-by-step recipe for an experiment. Then there’s plumbing work, too: Opentrons wants its robots and instruments to plug more deeply into the rest of the lab automation world.
What are autonomous lab robots, anyway?
Think of a lab bench where a robot handles the liquids, the plates and the timing. Now add an AI on top that decides what to test next. That’s the basic idea.
Opentrons describes a loop where its AI and robots “design, execute, and learn from scientific experiments.” So the AI plans a test, then the robot runs it. Next, the results feed back into the AI, and the cycle starts again. If you’ve ever wondered how a machine turns a command into a precise movement, our guide to robot actuators covers the basics.
CEO James Atwood put it plainly: “our systems are built for AI to run experiments directly.” He went further, too: “That is why autonomous science is standardizing on Opentrons.” Big claim. We’ll come back to it.
Where do the smarts come from?
Training a robot takes lots of data, and real lab time is expensive. So earlier in 2026, Opentrons said it uses NVIDIA’s Isaac and Cosmos platforms to create training datasets for its lab robots. In other words, both are NVIDIA tools for building and training robots. If you want the bigger picture on how AI makes sense of mixed inputs, our explainer on multimodal AI is a good start.
Why this one’s a big deal
Lots of robotics startups have slick demos and few customers. Opentrons isn’t one of them. It says it has more than 11,000 robots in labs across more than 80 countries. Its main platform, Opentrons Flex, pairs those robots with AI systems. And according to Robotics & Automation News, 90% of the world’s top 20 drug companies use its systems.
That install base matters, because AI can’t run experiments without robots already sitting in real labs. The company also brings “more than a decade of laboratory automation experience,” as the same outlet put it.
Avego’s managing partner, James Flexner, nailed the core argument: “Every AI-driven discovery still has to be tested in a real lab.” Honestly, that’s the line that sold us. All the clever models in the world don’t mean much until a liquid moves from one tube to another.
Honestly, is $40 million a lot?
Not by today’s robot standards. Humanoid startups are pulling in far bigger numbers. As we covered recently, Physical Intelligence could be worth about $20 billion without a dollar of revenue.
But that’s sort of the point. Lab robots aren’t glamorous. They don’t do backflips. Yet they already work, they’re already installed, and they already have paying customers. Most days of the week, we’d take a boring robot that ships over a shiny one that doesn’t.
Okay, but should you care?
Probably more than you think. That’s because faster lab work can mean faster drug discovery, quicker diagnostic tests and cheaper research. Those gains eventually reach patients, wherever they live.
There’s an angle for smaller countries too. Atwood says Opentrons “made lab automation accessible for any lab.” If that holds, a university lab in Yerevan or a young biotech startup won’t need a giant pharma budget to automate its work. And since the company plans to enter new markets and regions, more of these machines could show up outside the usual hubs.
It also changes what lab skills look like. Tomorrow’s lab tech might spend less time pipetting and more time checking what the AI wants to test next. We think that’s a shift worth preparing students for now.
Where autonomous lab robots go from here
So, is autonomous science really “standardizing” on one company? That’s the CEO’s pitch, and it’s way too early to call. We’d also expect plenty of others to chase the same prize.
Still, the direction feels right. Autonomous lab robots turn AI from a smart advisor into something that can actually test its own ideas. The real question is what happens when thousands of these loops run every night. Will we get better medicines faster, or just a mountain of data nobody has time to read?