Fully autonomous lab robots.
A mobile robot that walks into the lab you already have, runs your experiments around the clock, and never asks you to rebuild the room.
Most autonomous labs start by asking you to build a new one.
Millennium doesn't. Our robot moves through an ordinary wet lab on its own, reaches the instruments you already own, and does the work at your bench. The room stays yours.
- We never askStationary arms Plate tracks between stations A renovated lab
- ReachInstruments within the robot's reach of the aisle.
- FiducialsSmall visual markers at each station so the robot localizes precisely.
- BenchOne dedicated bench for liquid handling, with our fixtures and tools.
- ConsumablesThe same vials, plates, and tips from one experiment to the next.
- ChargingPin-and-prong contacts at key stations so the robot tops up while it works.
- BaseA corner of the lab for charging and calibration.
A lab tech that works every hour of the week.
A founder with a first check has to prove a target or a small-molecule program before the money runs out. Hiring techs is slow and expensive. Outsourcing to a CRO means someone else runs your science, at their pace and their margin, with your IP in their hands.
Millennium rents the robot. It runs your protocols overnight and through the weekend, logs every step, and pivots when the AI in the loop decides the next experiment. You keep the science, the data, and the feedback loop.
| Lab tech | Millennium | |
|---|---|---|
| Cost | $100k+ per year, plus benefits | $3–5k per month |
| Hours | 40 per week | 168 per week |
| Coverage | One person, daytime | Robot plus AI, day and night |
| Record | Notebook and memory | Every action logged and replayable |
Rental pricing is a target for the first commercial units.
The experiment doesn't wait for Monday.
The robot runs the protocol. The model reads the result, decides what to change, and the robot runs the next one. No overnight idle, no waiting for a scientist to get back to the bench.
Because a program executed every step, each run is repeatable and traceable. That's the difference between a result and a result you can trust.
Early-stage R&D is leaving the country.
Early biotech research costs three to five times less overseas, and it moves faster. With the same money, a founder abroad runs several programs in parallel and finds the one that works. In the US, cost, labor, and slow iteration force a single bet.
Robots that work around the clock in labs that already exist are the fastest way to give US founders the cost and speed back. The country has over 100 million square feet of wet lab space. None of it needs to be rebuilt.
When ideas outrun instruments.
As models generate more good hypotheses than any lab can test, the bottleneck moves to the bench, and then to the instruments themselves. We're building toward robots that can design, assemble, and iterate on the instruments the next experiment needs.
Global first place, WBCD 2026.
Millennium is built by Team Proxemics, global winners of the XtalPi Lab Experiments Track at What Bimanuals Can Do 2026, held at IEEE ICRA 2026.