Binding self-driving laboratories into a national network - AIMD-NET for materials design (NSF award $5M, $20M estimated total)
The U.S. National Science Foundation awarded Johns Hopkins University $5 million for research binding self-driving laboratories - robotic experimentation combined with distributed AI - into a national-scale network.
Grant overview (primary data)
- Award amount$5,000,000 / Est. total $20,000,000
- RecipientJohns Hopkins University (Maryland)
- ProgramPCL-Programmable Cloud Labs, PCL-Programmable Cloud Labs
- Period2026-08-01 〜 2030-07-31
- FunderU.S. National Science Foundation (NSF) / NSF
Key points
- A self-driving laboratory combines robotic experimentation with AI that decides what to try next, moving judgement to the machine.
- The subject of the award is extending autonomy from instruments and single laboratories to a national network.
- It names real-time cross-site exchange, a multi-agent decision fabric with shared uncertainty quantification, and AI-ready data generation.
- The target is accelerating structural materials development, including for extreme environments in defense, transportation and energy.
- Of the 120 NSF awards this site holds as of 2026-09-02, 11 fall under the same programme, totalling about $45.79 million across 11 institutions.
- The $5 million obligated sits against a $20 million estimated total, so funding proceeds in stages across years.
1What a self-driving laboratory is
A self-driving laboratory combines automated robotic experimentation with AI that decides what to try next. A person forms a hypothesis, sets up the experiment, reads the result and decides the next step; handing that loop to machines aims to raise the rate at which knowledge itself is produced. What separates it from ordinary automation is that the judgement, not only the execution, moves to the machine.
2From instrument to laboratory to network
What this award addresses is how far that autonomy can be extended. One autonomous instrument does not make the whole faster if it cannot communicate with the instrument beside it.
- 1Instrument levelA single instrument sets its own measurement conditions and returns results
- 2Laboratory levelSeveral instruments coordinate within one laboratory
- 3Network levelSites exchange information in real time and AI decides in coordination
The award names an event-driven, interoperable cyberinfrastructure for cross-site exchange in real time, a modular multi-agent decision fabric with a shared schema for quantifying uncertainty, and machinery producing data in a form AI can train on directly. Results are not read by a person deciding what comes next; they flow in a form machines can read.
3Eleven sites started on the same day
This is not a standalone project but one of eleven sites stood up simultaneously under the same programme. As the phrase test bed suggests, each site varies its subject field while testing a common foundation. Read alone it looks like a grant to one university; bound by programme name it is a national mesh being laid.
4The gap between obligated and estimated total
The obligated amount is $5 million against an estimated total of $20 million. Large NSF awards are funded in stages across years, so the first-year figure understates the scale. Of the 120 awards this site holds as of 2026-09-02, 28 carry an estimated total more than 1.5 times the amount obligated.
Why it matters
Materials development is costly in time and money because it proceeds by trial; handing the design and execution of experiments to machines compresses the development period itself. Autonomy across sites presupposes machine-readable data, a shared way of expressing uncertainty and real-time communication between sites. For firms in materials, chemicals and semiconductors, how such a common foundation is standardised bears on their own capital investment and data preparation.
FAQ
What is a self-driving laboratory?
Why does it need to be a network?
Why do the obligated amount and estimated total differ?
Sources (primary)
Source: NSF Award Search (U.S. National Science Foundation, public domain). Amounts are the obligated amount. For privacy, we do not handle principal investigator names.
- NSF Award (original, official)
- NSF Award ID: 2607526