top of page

Two Robots Have Crossed Into Decision-Making. Neither Is Available to a Surgeon.

Aug 26
3 min read

Updated: Aug 30

Two Robots Have Crossed Into Decision-Making. Neither Is Available to a Surgeon.
Credit: Juo-Tung Chen/Johns Hopkins University/PA

The surgical robotics industry uses the words AI, automated, and intelligent almost interchangeably.


Few of the systems that carry those words mean the same thing by them.


The Surgical Autonomy Registry sorts that language by a single variable: who makes the decision, and who executes the movement.


Nine systems, from a 1994 laparoscope holder to a research robot that has never touched a living patient, get placed on a six-level ladder and sourced to a regulatory filing, a peer-reviewed study, or an official company statement.


That laparoscope holder, AESOP, moved the camera on voice and foot-pedal command and made no decisions of its own. Source: Journal of Visceral Surgery


Almost everything on the market assists. Almost nothing decides.


Where the Market Actually Sits


Every soft-tissue robotic system currently sold lands at L1, Assisted: da Vinci Xi and da Vinci 5, Hugo RAS, Versius, and OTTAVA.


Da Vinci filters the surgeon's hand tremor and scales large movements into small instrument movements; the surgeon makes and executes every decision at the console. Source: Intuitive Surgical


Versius performs surgical tasks under the control of the surgeon and in response to input from the two hand controllers, per its own FDA summary. Source: FDA De Novo Summary


None of them makes a surgical decision on its own.

OTTAVA is the clearest case in the registry.


Its arms move into position at the push of a button, and Johnson & Johnson's own materials describe that as automated. Source: Johnson & Johnson MedTech


Positioning an arm before surgery starts is not the same as performing the surgery.


The incision, the dissection, and the suture all stay with the surgeon at the console.


Hugo's AI-powered language works the same way: it describes an analytics layer that reviews a case after it ends, not a system deciding anything during it. Source: Medtronic


The One Real Second Rung


Orthopedics is the only category that has genuinely moved past L1.


Stryker's Mako uses haptic technology to enforce a pre-planned virtual boundary; the surgeon still moves the saw or burr by hand. Source: Stryker


Zimmer Biomet's ROSA Knee positions and holds a cutting guide while the surgeon performs the bone cuts by hand, described in peer-reviewed literature as collaborative robotics in which the surgeon remains in charge. Source: Archives of Orthopaedic and Trauma Surgery


That is L2, Shared Control: a real function running alongside the surgeon, not replacing any part of the decision.


The Claim That Doesn't Match the Product


Monogram Technologies' mBôs TKA system is the registry's clearest case of a claim exceeding its evidence.


The company describes its FDA-cleared implant-insertion function as designed to autonomously execute optimized paths.


No independent, peer-reviewed study backs that characterization yet.


A separate, not-yet-cleared configuration, a hands-free, saw-based version that performs the bone resection itself, is the one actually approaching L3.


That version has not been demonstrated on a live patient and is not commercially available. Source: GlobeNewswire


Two different capabilities share one brand name, and any headline calling Monogram autonomous is describing one of two very different things.


The Two That Aren't Products


Only two systems in the registry have published evidence of completing a real surgical task independently.


Johns Hopkins' STAR robot autonomously completed more than 83% of a suturing task while reconnecting intestine in live pigs, and a human corrected the stitches it missed. Source: Science Robotics


SRT-H, built by Johns Hopkins and Stanford, autonomously performed the clipping-and-cutting phase of a gallbladder removal across eight ex vivo organs, with a 100% success rate. Source: Science Robotics


Neither system has an FDA filing. Neither is available in an operating room.

What This Actually Means


The ladder does the same job everywhere it's applied: it separates a system that assists a human decision from one that makes one.


As of this edition, two systems in the world have published evidence of making that decision.


Neither is for sale.



The full registry breaks down all nine systems, their regulatory status, and the primary source behind every placement: get The Autonomy Registry.


An interactive version of the ladder is live at nextinsurgery.com/surgical-autonomy-ladder.


Related reading

icon-autonomy.png

Autonomy

Four robots claim "autonomous." Here's where each one actually sits.

What's actually true about AI, autonomy, and surgical robotics?

Four reports. Every claim checked against FDA filings and peer-reviewed evidence — not press releases.

icon-ai-claims.png

AI

Five FDA-cleared systems. What's cleared, inferred, or just marketing.

The Evidence Library

icon-registry.png

Registry

Nine robots mapped across six levels of autonomy.

icon-roadmap.png

Roadmap

Intuitive's five-layer autonomy roadmap, audited layer by layer.

bottom of page