The robot does not lift anyone. It does not change a dressing or push a med cart. What it does, several hundred times a day inside American skilled nursing facilities, is roll up to a bed, read a resident's vital signs without touching them, run a structured assessment in whatever language the resident answers in, and write the encounter into the electronic health record.
That is the entire product, and on Sept. 1 investors put $14 million behind it.
Norbert Health, a Brooklyn company whose software turns third-party robots into what it calls autonomous nursing assistants, announced a Series A supported by William A. Marino with Cardinal Group, with participation from Exor Seeds and CareIT. The angel list leans operator-heavy: Datadog co-founders Alexis Le-Quoc and Olivier Pomel, Noom co-founder Saeju Jeong, and Owkin co-founder Thomas Clozel. Total raised to date: $19 million.
The pitch is a critique of everything that came before it.
“Healthcare in the US is short close to a million nurses, and more than two million trained nurses no longer want to work bedside because the job keeps getting harder,” said Alex Winter, Norbert's CEO and co-founder. “Robots that assist nurses can help reduce the burden, but only if they have clinical skills. Previous attempts put logistics robots on healthcare floors and never touched the core of a nurse's work.”
The Brain, Not the Body
Norbert does not build robots. It builds the control system that rides on someone else's hardware — the company's framing is that machines that can already walk, navigate and carry are missing the clinical brain and voice that would make them useful at a bedside.
That stack has two halves. The sensing side covers contactless physiology, mobility and motor function, cognitive and mental-health markers, and behavioral and social activity. The interaction side handles conversation across languages and communication styles, runs structured assessments, and coordinates with staff. Together, Norbert argues, they catch the small deteriorations that turn into readmissions.
The deployment record is the reason this round happened. Norbert has been live in skilled nursing facilities since August 2025, rounding on hundreds of patients daily. Patient acceptance is 96%. Monitoring compliance has reached 82%, which the company says is nearly twice the baseline for manual remote monitoring — a claim that, if it holds up under independent scrutiny, is the commercially interesting number in the announcement. Compliance is where remote-monitoring programs usually die.
The new capital goes toward more clinical skills, regulatory clearances, and deployments beyond skilled nursing. In development: post-fall neurological evaluations, pressure-injury prevention protocols, cognitive screening, and fall-risk reassessment. Each ships on top of the installed base, which is the whole economic argument — sell the robot once, then keep adding billable capability by software update.
The Money Around It
Norbert's $14 million is small change in the current robotics market. Crunchbase counts $18.8 billion into robotics startups globally so far in 2026, against $15 billion for all of 2025 and $14.1 billion in the previous peak year of 2021. Individual rounds dwarf this one: Saronic took $1.75 billion, Skild AI $1.4 billion.
But those are bets on building machines. Norbert is a bet that the machines will be commodities and the skills will not.
“Norbert is the rare physical AI company already earning its place in live clinical operations: real patients, real facilities, real workflows, every day,” said Filippo Monteleone, President and Founder of CareIT. “The deployments are working, and the lead grows wider from here.”
Jeong, the Noom co-founder, framed it as a market-access story: “Healthcare labor is one of the largest markets in the US economy, and robots have never had a credible route into it until now.”
Why This Matters
Three things are converging here, and only one of them is about robots.
The first is that physical AI is finally producing revenue rather than demo reels. Most of the sector's 2026 funding is still underwriting capability that has not met a paying customer. Norbert's differentiator is a year of production hours in a setting — post-acute care — that is chronically understaffed, heavily regulated, and financially squeezed enough to actually try something new.
The second is the labor math. Skilled nursing facilities are the part of American healthcare where the staffing crisis is least abstract. Winter's framing is careful and worth noting: not replacement, but the removal of repetitive monitoring work from people whose scarce time should go to judgment and hands-on care. Whether operators buy it for that reason or as a headcount substitute is a different question, and the answer will be set by facility economics rather than by Norbert.
The third, and least resolved, is regulatory. The company says the round will help it advance regulatory clearances, which is a straightforward acknowledgment that it does not have all of them yet. Contactless vital-sign capture and structured clinical assessment sit close to the line where software becomes a regulated medical device, and the roadmap pushes harder against it: a post-fall neurological evaluation is not documentation, it is triage. If an autonomous system runs that assessment and misses something, the accountability chain — vendor, hardware partner, facility, supervising nurse — has no settled answer.
Consent is the quieter problem. A 96% acceptance rate is a strong number, but it is measured in a population with significant cognitive impairment, and the standard for informed agreement to being continuously sensed by a robot is not obvious.
What to Watch
Which clearances Norbert actually pursues, and for which skills — that will reveal how aggressively it reads its own regulatory exposure. Whether the ship-to-installed-base model holds once capabilities get clinically heavier. Whether the reimbursement pathways the company says its robots help activate survive contact with payers at scale. And whether anyone publishes outcomes data: the case for autonomous rounding rests on catching acute events early, and so far that claim exists only in the company's own telling.
“Robots that assist nurses can help reduce the burden, but only if they have clinical skills. Previous attempts put logistics robots on healthcare floors and never touched the core of a nurse's work.”— Alex Winter, CEO and Co-Founder, Norbert Health