Atinary brings Self-Driving Labs to Boston: AI that runs real experiments
Date: 11 Feb 2026
Atinary has opened a new AI-powered Self-Driving Lab in Boston, pushing AI beyond prediction and into the physical execution of science. The goal is straightforward: faster, more reliable discovery across chemistry, materials, and pharmaceutical R&D.
The facility runs two autonomous platforms-Atinary's Scientific Discovery Factories-built to continuously design, execute, analyze, and learn from real experiments. Every experiment feeds back into Atinary's machine-learning stack and foundation model, compounding insight with each iteration.
Why this matters for your R&D
- Closed-loop Design-Make-Test-Analyze-Learn (DMTAL) cycles that reduce manual handoffs and wait time.
- Higher reproducibility and data quality through standardized execution and integrated analytics.
- Faster convergence on optimal chemistries, catalysts, and process conditions.
- More efficient use of reagents, instruments, and scientist time.
How the platform works
Atinary integrates its no-code AI platform with robotics and instruments from ABB, Agilent, Bruker, Chemspeed, and METTLER TOLEDO. Experiments no longer run as single-threaded, manual sequences-They execute as closed loops where outcomes automatically determine the next best set of experiments.
The system connects planning, execution, analytics, and learning natively. That means better data capture by default, fewer reruns, and a tighter link between hypotheses and validated results.
What's in scope today
The Boston lab focuses first on small-molecule synthesis and catalysis. It supports pharmaceutical R&D from early discovery through process development, and the platform is built to scale across broader chemistry and materials programs.
What stays human
This model augments scientists rather than replacing them. Human judgment sets goals, constraints, and quality thresholds, while machine-speed execution explores the search space and flags what to test next.
Work is guided by Atinary's multidisciplinary team and Scientific Advisory Board spanning chemistry, AI, supercomputing, and lab automation-including MIT Professor Stephen Buchwald.
What changes in your workflow
- Move from static design-of-experiments to adaptive, AI-guided campaigns.
- Shift from fragmented software and manual exports to an integrated data backbone.
- Treat every experiment as both a result and a learning event that improves the next run.
The lab as a production-grade platform
Atinary delivers the Self-Driving Lab as a fully integrated stack: human intelligence, artificial intelligence, robotics and automation, and precision instrumentation. The outcome is a frictionless R&D environment where execution, analytics, and learning are connected by default.
Further reading and upskilling
For background on autonomous discovery, see this review on autonomous discovery in chemistry.
If your team is building skills in AI-assisted lab automation and closed-loop experimentation, explore curated training paths: AI courses by job.
Your membership also unlocks:
