NSF AI award $5M: a cloud lab node for materials and manufacturing (Georgia Tech) — ten nodes launched the same day, $185.7M in planned totals
NSF awarded $5,000,000 to Georgia Tech Research Corporation for "PCL-Test Bed: The AI Materials and Manufacturing Pilot Facility Node (PCL-AIMMPF)." The estimated total is $18,150,000, running from August 2026 to July 2030. Looking across our collected data, ten of these Programmable Cloud Lab test beds all begin on the same day, with $46.1M obligated and $185.7M in estimated totals.
Grant overview (primary data)
- Award amount$5,000,000 / Est. total $18,150,000
- RecipientGeorgia Tech Research Corporation (Georgia)
- ProgramPCL-Programmable Cloud Labs, PCL-Programmable Cloud Labs
- Period2026-08-01 〜 2030-07-31
- FunderU.S. National Science Foundation (NSF) / NSF
Key points
- NSF Award 2607581, "PCL-Test Bed: The AI Materials and Manufacturing Pilot Facility Node (PCL-AIMMPF)," to Georgia Tech Research Corporation (GA).
- Program: PCL-Programmable Cloud Labs. $5,000,000 obligated; estimated total $18,150,000.
- Period 2026-08-01 to 2030-07-31.
- Within our collection there are ten "PCL-Test Bed" awards sharing identical start and end dates, totaling $46.1M obligated and $185.7M estimated, across seven states.
- A detailed abstract of the research is not included in this data.
- The ten awards titled PCL-Test Bed share start and end dates, totalling $46.1M obligated and $185.7M estimated.
1What a "cloud lab" actually refers to
Programmable Cloud Labs (PCL), the name of the program, refers to research facilities whose instruments can be called remotely. Rather than operating equipment by hand, a user writes the experimental procedure as a program and sends it; the instruments execute it automatically and return the resulting data. It is the same idea as renting computing capacity in the cloud, applied to physical experiments.
Arranged this way, the experiment itself turns into machine-readable procedure and machine-readable data. Conditions and results that used to be written in a notebook become material an AI can learn from directly. That, more than raw instrument speed, is why automated experimentation is drawing attention.
2A "pilot facility" between the bench and the factory
Pilot Facility, in this award title, is the term for a mid-scale setup that bridges small benchtop preparation and factory-scale production. In materials work it is common for a substance that shows good properties in the lab to stop showing them once the quantity increases. Providing the equipment that crosses that gap — and making it usable remotely — is how this node reads.
3Ten nodes starting on the same day
In our collected data, ten awards carry "PCL-Test Bed" in the title, and every one of them starts on 2026-08-01 and ends on 2030-07-31. Together they represent $46.1M obligated and $185.7M in estimated totals, with recipients across seven states. That pattern is hard to notice when reading award records one at a time, but it says this is not scattered project support: it is a nationwide network being stood up at once.
Why it matters
Once experiments can be executed remotely as procedures, conditions and results accumulate in a machine-readable form from the start. The constraint on AI-driven materials search is usually the volume and quality of data rather than the algorithm, so this kind of capital investment shapes search speed itself. That nodes of one program stand up in seven states at once also points to a design choice: keep the instruments regionally distributed while making them usable through common procedures.
FAQ
What is a Programmable Cloud Lab?
Why do the obligated amount and the estimated total differ?
Is "ten nodes" a figure published by NSF?
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: 2607581