Automation vs Autonomy in Surgical Robots
A surgical robot can perform a task automatically without being autonomous.
The Simple Difference
Automation means the surgeon determines the task, and the robot carries it out.
Autonomy means the robot helps decide what to do, based on the patient or the situation.
The key distinction is where decision-making sits: with automation, the surgeon sets the action; with autonomy, the system takes on part of that decision-making.
A systematic review of 49 FDA-cleared surgical robots in npj Digital Medicine gives a clear way to sort them.
The Three Levels That Exist Today
Level 1, Robot Assistance: the surgeon controls every movement, and the robot only makes those movements steadier or more precise.
The da Vinci Surgical System is a familiar example.
Level 2, Task Autonomy: the surgeon sets up one task, and the robot performs it without continuous hand control.
This is automation, because the robot executes a plan the surgeon already made.
Level 3, Conditional Autonomy: the robot builds a patient-specific plan, the surgeon approves or edits it, and the robot carries it out.
The TSolution One is an example, since it creates a surgical plan from the patient's anatomy and can then mill the bone automatically.
Level 4 (High Autonomy) and Level 5 (Full Autonomy) would involve the robot choosing the strategy and eventually running whole procedures, and no cleared system in the review reached either one.
Where FDA-Cleared Robots Actually Stand
The review covered 49 unique surgical robots cleared by the FDA from 2015 through 2023.
42 systems (86%) were Level 1.
4 systems (8%) were Level 2.
3 systems (6%) were Level 3.
0 systems were Level 4 or Level 5.
Most cleared surgical robots are still tools that the surgeon controls directly.
What Research Robots Are Starting to Do
The review only counts robots the FDA had cleared by December 2023, so research systems fall outside its numbers.
One example is SRT-H from Johns Hopkins, which completed the clipping and cutting phase of a gallbladder removal on eight ex vivo pig gallbladders without human intervention.
That result shows a robot deciding its next step within a procedure, which is the kind of capability the higher levels describe.
It is still a laboratory result on tissue outside the body, so it is not a cleared product, and it does not move any FDA-cleared robot above Level 3.
What to Check Before You Trust a Label
Clinical evidence is thin, because only 19 of the 49 systems (39%) came with clinical testing data.
All three Level 3 systems had it.
Marketing language also does not always match the FDA record, since three systems were advertised as machine-learning-enabled on company websites while their FDA summaries did not list those features.
The practical rule is to trust what the FDA submission says the system does, not what the product page calls it.
Three Questions to Ask
When a robot is described as automated or autonomous, ask these in order:
Does the surgeon control every movement? If yes, it is Level 1.
Can the system perform a task without continuous control? If yes, it is at least Level 2.
Can the system create a patient-specific strategy on its own? If yes, it is Level 3 or higher.
The progression is Assistance, then Automation, then Autonomy.
Assistance changes how the surgeon acts, automation changes who executes the task, and autonomy changes who decides the strategy.
For the full comparison: Surgical Autonomy Registry
