Self-Driving Labs Draw US$575M in 17 Public Awards While No Private Round Funds an Autonomous Laboratory as a Product in Six Months of Corporate Announcements
A research-led analysis of 78 research and 26 corporate records finds that shared and cloud laboratory capacity rose from 4 to 22 percent of research between halves, while nobody is selling the connective layer between public capacity and private demand.

InnoDexis has published its latest Innovation Intelligence Report covering self-driving laboratories and autonomous experimentation, analyzing 78 research records from 49 institutions in 12 countries between 15 October 2025 and 9 October 2026, and 26 corporate records between 13 April and 2 October 2026. The report reveals that the laboratory layer is being built with public money and almost no private capital: US$575M across 17 public and philanthropic awards funds autonomous laboratory capacity, while no private round for an autonomous laboratory as a product is visible.
Key Findings
Shared capacity is the fastest-moving part of the field. Cloud laboratories, user facilities and consortia rose from 4 percent to 22 percent of research records between halves, a 17.5-point gain, while robotic hardware lost 15.9 points, and research volume rose from 3.8 records a month to 8.7.
The money is public. US$575M across 17 awards is led by Toronto’s US$200M Acceleration Consortium and the NSF’s US$108M across six materials centres, alongside NSF cloud-laboratory nodes and Germany’s €30M ASCEND consortium, while the only private round in the window, Fathom Therapeutics’ US$47M Series A, funds a drug-discovery company that uses the method.
Measured gains are large. Argonne ran more than 6,000 battery experiments in five months against a stated five to eight years conventionally, NC State found best-in-class nanoplatelets in 120 experiments over 12 hours, and the Fritz Haber Institute beat an industry-reference catalyst in fewer than 50 experiments from more than 10¹³ options.
The field is earlier than its headlines suggest. Only 12 of 78 research records describe a laboratory actually operating, 33 are concepts or roadmaps, and human expertise, workflow and adoption is the most-cited barrier, named in 17 of 72 records, ahead of reproducibility and validation at 16 and hardware and integration at 14.
Biology and data lead the commercial edge. Stanford tested 384 protein variants in about four hours of hands-on work, roughly 50 times faster than cloning, and autonomous data is the only thread where corporate share, 23 percent, exceeds research share, 14 percent.
Strategic Insight and Trend Analysis
The defining structural finding is that the capacity is public, the demand is private and nobody is selling the connection. Seventeen awards are building autonomous laboratories at NSF test beds, DOE national laboratories, Toronto, Berlin and Georgia Tech, whose cloud laboratory is designed to connect 150 institutions and serve more than 400 users, and Northwestern’s NSF-funded cloud laboratory will synthesise over 300,000 proteins. Yet no company in the dataset sells an autonomous laboratory, cross-site orchestration, scheduling or access brokerage.
The facility is replacing the instrument. Build versus buy has become build versus apply, since the capacity is deliberately multi-tenant, and industry enters through consortia, with BASF named six times and Siemens Energy four through ASCEND, rather than as customers. Northwestern argues that megalibraries screening a million samples in under 30 minutes outpace serial self-driving labs, so the architecture should be chosen by cost per experiment.
Several limits apply. The corporate window starts on 13 April, so any round announced earlier is invisible, the corporate base of 26 is illustration rather than a market measure, the report was assembled from records screened for two earlier editions, and the maturity ladder is rule-based, so the finding is that no round is visible in six months of English-language announcements.
Global and Industry Implications
For corporates and R&D teams, apply to a facility rather than budgeting a build, since NSF test beds, DOE user facilities and Toronto’s consortium take outside users, and pick one campaign with a large design space and an expensive experiment. Budget for scientific staff who can design campaigns before equipment, and confirm who owns the data generated under any agreement.
For investors and capital allocators, the absence of private rounds is a timing signal rather than a verdict, and the open position is the connective layer: cross-site orchestration, scheduling, access brokerage and data standards, with research proof at Carnegie Mellon, PNNL and UChicago. Data products built on public capacity, such as ADME-One and Acro-AIx, are where paying customers already exist, and a full twelve-month corporate window should be checked before acting.
For policymakers and national innovation bodies, access terms will decide whether US$575M creates a competitive market or closed campuses, so publishing access routes, pricing and data-ownership rules is the first lever. Twenty NSF test beds are being built in parallel with no claimed interoperability standard or reproducibility protocol, and requiring data-format compliance in new awards is the second.
InnoDexis Statement
“The capacity is public, the demand is private, and nobody is selling the connection — the gap is not capability or demand but the commercial layer between them,” noted InnoDexis in its latest intelligence report.
Conclusion
Because the corporate stream covers only 13 April to 2 October 2026 and the corporate base of 26 records is illustration rather than a market measure, the findings are cross-sectional and the absence of private rounds is a narrower claim than none exists. The signals to watch are whether NSF cloud-laboratory test beds publish external access terms, whether a private round funds an autonomous-laboratory or orchestration product, and whether cross-site orchestration is released as software. The complete Self-Driving Labs & Autonomous Experimentation Report October 2026 is available to InnoDexis subscribers and enterprise clients.
About InnoDexis
InnoDexis is a real-time innovation intelligence platform connecting scientific discovery to commercial opportunity. Powered by InnoGraph, its proprietary innovation knowledge graph, the platform maps technologies across the entire lifecycle, from early-stage research (TRL 1) to market deployment (TRL 9), connecting researchers, institutions, startups, companies, products and emerging opportunities.
Covering 40+ specialised industry segments and 360 technology areas, InnoDexis delivers innovation tracking, research discovery, product and startup intelligence, investment insights, and AI-powered reports. Its mission is to bridge the gap between scientific discovery and real-world impact.
InnoDexis — From Lab to Market: Real-Time Innovation Intelligence.