Every Surgical Robot Claims Autonomy. Only Six Levels of It Exist.
Updated: Aug 30
A scalpel never decides anything; it waits for a hand to move it.
The moment a robot starts deciding instead of waiting, the words describing it need to be precise, and in most marketing copy, they aren't.
Where the Framework Comes From
In 2017, a group of robotics researchers published a paper in Science Robotics proposing six levels of autonomy for medical robots, adapted loosely from how the auto industry classifies self-driving cars.
That six-level scale is the backbone of the Autonomy Ladder used on this site.
The Six Levels, Plainly
Level 0 is no autonomy: the robot provides no independent motion at all; it's a tool a surgeon holds and moves directly.
Level 1 is robotic assistance: the system helps the surgeon's hand, like filtering out tremor, without making any decisions of its own.
Level 2 is task autonomy: the robot can execute a small, well-defined action, like stopping a saw at a preset boundary, under continuous human control.
Level 3 is conditional autonomy: the robot performs a specific surgical subtask on its own, like a suturing sequence, while a surgeon supervises and can intervene at any moment.
Level 4 is high autonomy: the system makes and executes a broader set of decisions under human supervision, without being directed step by step.
Level 5 is full autonomy: the robot decides and executes without any human supervision, a level that does not yet exist in a real surgical device.
The Marketing Words vs. the Ladder
"AI-powered" almost always describes Level 0 or Level 1, decision support or image analysis that a surgeon still acts on manually.
"Autonomous" is the most overused word on this ladder, applied loosely to everything from a robot executing one preplanned cut, which is closer to Level 2 or 3, all the way up to genuine claims of Level 4.
"Semi-autonomous" is usually an honest label, typically describing real Level 2 or Level 3 behavior.
"Closed-loop" is a technical term, meaning the system senses and reacts in real time, and it can appear at almost any level, so it says nothing about how much of the decision-making the robot actually owns.
How to Read Any Claim
A precise autonomy claim answers three questions: who makes the decision, for what specific subtask, and under what level of supervision.
If a claim can't answer those three questions, it isn't describing a level of autonomy; it's describing a feeling.
Source: Science Robotics
The Autonomy Claims Audit checks every major manufacturer's autonomy language against its own FDA filings, claim by claim.
