Lab automation

Agents run experiments, not only propose them.

Facts

Lane: Science and health
Wave: 3 of 12 (build order, not a date)
Needs: 2 boxes all the way back through hard links
Unlocks: 5 boxes downstream through hard links
Leverage: 0.4 points · stands in the way of 2 of 53 stages · rank 20 of 57
Cluster: Science and medicine

Evidence as of October 2026

FutureHouse’s Robin proposes experiments and interprets the results in a closed loop, while people still run the wet-lab work. Fully automated labs are the next step.

01Lab in the loop: from research tool to research leader

Around it

Needs · 2

Robot data and foundation modelscapability · Emerging
World models and simulationcapability · Emerging · helps, not required

Unlocks · 5

Perturbation data at scalecapability · Emerging
Materials discoverycapability · Emerging
AI-designed medicinesstage · Emerging
Agent-run discoverystage · Emerging · helps, not required
Experimental validation capacitycapability · Emerging · helps, not required

Loops and off-ramps

Science feeds the substratereinforcing loop
Capex pullbackoff-ramp · slowed
Taiwan chip disruptionoff-ramp · slowed
Power shortfalloff-ramp · slowed
Training data ruled outoff-ramp · slowed

Openings researchers' judgement, October 2026 scouting

Startup room 42/100 · rank 56 of 124 Research room 66/100 · rank 11 of 124

Crowding3 Lab overlap3 Small-team fit2.5 Buyer today2.5 Research openness5

Depth: deep
Players: Lila Sciences ($550m, $8.5bn talks); Periodic Labs ($300m seed); Edison Scientific ($70m); Emerald Cloud Lab; Toronto SDLs
Wedge: Closed-loop experiment selection and scheduling software (Bayesian optimisation, metaheuristics) under noisy, costly evaluations, plus independent verification of AI-proposed results. research