FDA posted a discussion paper on November 6 asking how it should weigh benefit and risk in robotic devices that carry some autonomy, are driven from a distance, or both. Comments go to docket FDA-2026-N-13135 through January 5, 2027. CDRH holds a hybrid public workshop on the same material December 2 and 3.
The paper covers four areas: terminology, benefit and risk for autonomous capabilities, benefit and risk for remote teleoperation, and risk mitigations. It walks through existing autonomy scales, including the six-level Yang framework and the degree-of-autonomy approach in IEC/TR 60601-4-1, then asks whether a single whole-procedure rating works when a system automates one task and leaves the clinician in charge of everything else.
The teleoperation questions are the sharper set. FDA asks about motion and video latency drifting out of sync, what happens when connectivity drops mid-case, whether force and velocity limits substitute for the tissue resistance a remote operator cannot feel, and how a sponsor tests resilience to a cyber breach that degrades the video feed. It also asks who stands by at the bedside, and notes that the sites gaining most from remote procedures tend to have the fewest local specialists.
Nothing here is guidance, and FDA says so twice in the first page. The 15 questions still read like the questions a reviewer will put to a surgical robotics startup in a pre-submission two years from now. Answering them inside your own hazard analysis now is cheaper than answering them in a deficiency letter.