Bleu Robotics raised a $5.4M pre-Seed round to cut humanoid robot deployment to under an hour. Its abductive learning model trains robots from just ten demos.
Bleu Robotics just raised a €5 million pre-Seed round — that's roughly $5.4 million — and the Paris-based team has a goal that sounds almost too bold: cut the entire humanoid robot deployment process down to an hour or less.
If you've followed factory automation at all, you know that's not how things usually go. Getting a robot ready for real production work can take weeks or months. Bleu thinks it can happen before your lunch break.
### The Problem With How Robots Learn Today
Here's the thing most people miss about humanoid robots. The hardware has gotten genuinely impressive. The software? Not so much.
Interact Analysis found that of the 20,000-plus humanoids produced last year, only about 10% actually reached real-world applications. That's a lot of expensive hardware sitting around waiting for software to catch up.
Most robots today rely on end-to-end models — basically a black box that learns everything at once and needs massive amounts of data to work reliably. Hundreds of hours of training data, and still no guarantee the robot performs on the line.
Benoît Berkoukchi, co-founder of Bleu Robotics, puts it plainly: "Most robotics companies are trying to train humanoids with millions of data points and build fully general-purpose models. That approach isn't suited to manufacturing."
He's got a point. There just isn't internet-scale data for factory work the way there is for language models. And there won't be for a long time.
### Doing More With Less
Bleu's answer is something called abductive learning. Instead of feeding a robot endless examples, the system lets it pick up a new task from as few as ten operator demonstrations.
That's the whole pitch. Ten demos, and the robot can get to work — while keeping the explainability industrial safety requires and the 99%+ accuracy production lines demand.
> "Our methodology means we can do more with less." — Benoît Berkoukchi, co-founder of Bleu Robotics
The software runs on any humanoid platform and bundles the AI model with the training devices needed to teach robots right on the factory floor. No lab required.
### Who's Backing This
The round was co-led by Contrarian Ventures and OVNI Capital, with participation from Kvanted, WEPA Ventures, Motier Ventures, Charlie Songhurst, and board members at Meta.
Thomas Poitrineau, an investor at Contrarian Ventures, explained the appeal: "It's rare to find a company led by two world-class robotics researchers and a manufacturing expert." He added that Bleu's approach "can already scale, putting robots to work in factories within months rather than years."
That mix matters. Bleu was founded in Paris in 2025 by two INRIA researchers and a former Renault manufacturing manager — deep research brains paired with someone who actually knows what a factory floor looks like.
### A Hot Corner of European Tech
Bleu isn't raising in a vacuum. European investment in robotics software, humanoid systems, and Physical AI has been surging through 2026.
- **Inbolt** (Paris) raised €11 million for its hardware-agnostic robot intelligence platform
- **Trener Robotics** (Norway and US) secured €26 million for AI-powered industrial robot skills
- **Motion** (Belgium) picked up €1.7 million for humanoid deployment
- **THEKER** (Spain) landed €73 million to scale generalist industrial robotics
- **Humanoid** (London) raised €133 million for its industrial humanoid platform
Add in rounds for Germany's Sereact, Switzerland's Enact Intelligence, and Poland's Nomagic, and you're looking at about €372.7 million in 2026 — around $403 million — climbing to roughly €377.7 million with Bleu's latest round.
### Why It Matters Beyond Europe
There were 5 million industrial robots operating globally in 2025, according to the International Federation of Robotics. But those machines repeat fixed tasks. They can't handle the high-mix, low-volume production that's now standard in modern manufacturing — so that work still lands on human workers.
Humanoids could change that. And Berkoukchi sees a bigger picture: "Humanoid robots in manufacturing are the next automotive sector, with the potential to help Europe and the US reshore production."
If Bleu can really get a robot trained in under an hour, that reshoring conversation gets a lot more interesting. For now, it's a $5.4 million bet that less data, not more, is the way forward.