ARPA-H Wants to Put a Robot Surgeon in Every Hospital. Here Is How to Get Funded.

Only 12 percent of Americans who need a thrombectomy — the procedure that removes a blood clot from the brain during an ischemic stroke — actually receive one. The other 88 percent either live too far from a hospital with a trained neurointerventionalist, arrive too late because of transfer delays between facilities, or never get diagnosed in time. Every ten minutes of delay in treatment costs roughly $10,000 in additional healthcare expenses and significantly increases the probability of permanent disability or death.

ARPA-H thinks robots can close that gap. Not robots assisting surgeons, not teleoperated devices controlled remotely by specialists — fully autonomous robotic systems that can perform an entire endovascular intervention without direct human input, stationed at any hospital in the country.

The agency's Autonomous Interventions and Robotics program, known as AIR, launched in November 2025 and represents one of the most technically audacious bets in the history of federal health research. As Granted News reported, full proposals are due March 18, 2026 — and the competitive landscape is already forming.

The Access Crisis That Created AIR

Stroke is the fifth leading cause of death in the United States and the leading cause of serious long-term disability. Approximately 795,000 Americans suffer a stroke each year, and roughly 87 percent of those are ischemic — caused by a blood clot blocking blood flow to the brain. For large vessel occlusions, mechanical thrombectomy is the gold-standard treatment: a catheter-based procedure where a physician threads a wire through the vasculature to physically extract the clot.

The problem is workforce and geography. There are fewer than 2,000 neurointerventionalists in the United States, concentrated in major metropolitan medical centers. The average American lives more than an hour from a thrombectomy-capable hospital. Rural communities — particularly across the Great Plains, Appalachia, and the Deep South — have almost no access. Patients in these regions routinely undergo inter-hospital transfers that consume critical treatment windows. Brain tissue dies at a rate of roughly 1.9 million neurons per minute during a stroke. By the time many patients reach a specialist, the therapeutic window has closed.

This is not a problem that can be solved by training more surgeons. The fellowship pipeline for neurointerventional surgery produces approximately 100 new specialists per year. At that rate, it would take decades to achieve anything approaching national coverage — and geographic distribution would remain skewed toward wealthy urban centers.

What AIR Actually Funds

The program is structured around two technical areas, each representing a fundamentally different approach to autonomous surgical intervention.

Technical Area 1: Autonomous Endovascular Robotics. This is the stroke-focused track. ARPA-H is seeking teams that can build robotic systems capable of performing a complete mechanical thrombectomy — from catheter insertion through clot retrieval — without a neurointerventionalist in the room. The system must handle real-time imaging interpretation, autonomous navigation through the cerebral vasculature, and adaptive decision-making when it encounters anatomical variations or procedural complications.

The vision is a self-contained unit that could be deployed at a community hospital, operated by existing nursing and radiology staff, and activated when a stroke patient arrives. The robot handles the procedure. The local team handles patient preparation and monitoring.

Technical Area 2: Interventional Microbots. This is the longer-horizon track. ARPA-H wants teams developing very small — millimeter-scale or smaller — autonomous devices that can navigate inside the body, perform targeted interventions, and exit or dissolve when the procedure is complete. Applications span oncology (targeted tumor ablation), urology (kidney stone removal), fertility treatments, and tissue biopsies. If Technical Area 1 is about putting robot surgeons in every hospital, Technical Area 2 is about making some surgeries obsolete entirely.

How the Money Works

AIR uses Other Transaction Agreements rather than traditional federal contracts or grants. OTAs offer significant flexibility: fewer regulatory constraints, faster negotiation timelines, and the ability to work with non-traditional defense and health contractors that might be ineligible for standard federal awards.

The program runs for five years, divided into a two-year Phase 1 and a three-year Phase 2. ARPA-H has not disclosed a total budget envelope — the agency rarely does for programs of this scale — but comparable ARPA-H initiatives like ADVOCATE (agentic AI for cardiovascular care) and UPGRADE (universal organ transplant) suggest individual awards in the range of $10 million to $50 million per performer, with total program investments potentially exceeding $200 million.

Critically, ARPA-H operates on milestone-based funding. Teams that fail to hit technical milestones at designated checkpoints will see their funding terminated. Teams that exceed expectations can receive accelerated follow-on investment. This is not an R01 where you write a five-year plan and check in at progress reports. The agency expects rapid iteration, and its program managers have the authority to restructure or cancel performer agreements at any checkpoint.

Who Should Apply

ARPA-H explicitly invites proposals from a broad coalition of organizations: private businesses (including startups), academic and scientific research institutions, hospitals, and hybrid teams that span multiple sectors. The agency anticipates — and actively encourages — teaming arrangements that combine robotics engineering with clinical expertise, regulatory strategy, and manufacturing capability.

If you are a robotics company or academic lab with strong endovascular navigation capabilities but no clinical trial infrastructure, find a hospital partner. If you are a medical center with stroke treatment expertise but no robotics engineering capacity, find a technology partner. ARPA-H's teaming profiles page is designed to facilitate exactly these connections.

Small businesses have a particular advantage under OTAs. Unlike traditional federal procurement, OTAs do not require DCAA-audited accounting systems or compliance with the full Federal Acquisition Regulation. Startups that would be excluded from DoD or NIH contracts can compete on equal footing.

What Reviewers Will Look For

ARPA-H evaluation criteria differ substantially from NIH or NSF peer review. The agency does not use study sections. Program managers — in this case, the AIR program director — have broad authority to select performers based on technical merit, team composition, and potential for transformational impact.

Based on ARPA-H's stated priorities and precedent from other programs, applicants should focus on several key areas. First, autonomous decision-making under uncertainty — the proposal must demonstrate how the system handles edge cases, anatomical variations, and unexpected complications without human intervention. A robot that works perfectly in simulation but requires a specialist to step in for 5 percent of procedures has not solved the access problem.

Second, regulatory pathway clarity. Any device performing autonomous surgery will require FDA authorization, likely through the De Novo or PMA pathway. Teams that have already engaged with FDA, obtained pre-submission feedback, or designed their clinical validation plan around regulatory requirements will be significantly more competitive than those treating regulatory strategy as a Phase 2 problem.

Third, deployment realism. ARPA-H wants systems that work in community hospitals — facilities with limited imaging suites, smaller nursing staffs, and no on-site specialists. A system that requires a $5 million angiography suite and three trained technicians to operate has missed the point. The proposal should address how the technology integrates with existing hospital infrastructure.

The Competitive Landscape

Several organizations have disclosed their participation in the AIR teaming process. Academic medical centers with strong neurointerventional programs — institutions like Mount Sinai, the University of Michigan, and UCLA — are natural clinical partners. On the robotics side, companies including Corindus (a Siemens Healthineers subsidiary that makes the CorPath robotic system for percutaneous coronary intervention) and several well-funded startups in surgical robotics are likely competitors.

The microbot track (Technical Area 2) draws from a different ecosystem entirely: MEMS labs, materials science groups working on biodegradable electronics, and several academic teams that have demonstrated proof-of-concept millimeter-scale locomotive devices in animal models. This is earlier-stage work, and ARPA-H is clearly comfortable funding high-risk proposals here.

Deadlines and Next Steps

Full proposals are due March 18, 2026, at 5:00 PM ET, submitted through the ARPA-H Solution Submission Portal. There is no separate letter of intent or white paper phase — the agency goes directly to full proposals under the ISO mechanism.

Applicants should review the AIR FAQ page carefully. Key administrative requirements include cost proposals with detailed milestone breakdowns, team composition narratives, and intellectual property plans. ARPA-H's IP framework under OTAs is more flexible than standard federal grants — performers typically retain ownership of foreground IP with the government receiving a license for government purposes.

For organizations considering a proposal, the March 18 deadline is tight. But ARPA-H ISO solicitations typically remain open for follow-on submissions even after the initial deadline, meaning strong proposals submitted later may still be considered. The first round of selections, however, will set the competitive pace for the entire five-year program.

If you have the technical capability to contribute to autonomous surgical robotics — in endovascular navigation, medical imaging AI, microelectromechanical systems, or clinical trial design — this is one of the largest, most consequential open funding opportunities in federal health research today. Tools like Granted can help you identify whether your team's capabilities align with ARPA-H's technical requirements and connect you with complementary partners before the deadline arrives.