What Humanoid Surgical Robots Would Need to Beat da Vinci

What UC San Diego's Humanoid Surgery Trial Showed

Researchers at the University of California San Diego have used a commercial humanoid robot to perform live surgery by teleoperation. The result, published in Nature in July 2026, is the first real test of a claim now circulating in surgical robotics. General-purpose humanoid surgical robots could eventually reach patients that a fixed system like Intuitive Surgical's da Vinci cannot.

The robot, nicknamed Surgie, is built on a Unitree G1 humanoid platform adapted with commercial laparoscopic instruments. Standing about five feet tall and weighing roughly 60 pounds, the robot completed two live laparoscopic gallbladder removals in animal trials. One procedure paired it with a human surgeon assisting. The second ran two humanoid robots working side by side. Neither required conversion to open surgery, though the team managed mild bleeding and bile leakage during both operations.

A da Vinci patient-side cart weighs close to 1,800 pounds and typically needs a dedicated, retrofitted operating room. Surgie's hardware fits in a single case and, its developers argue, could work in an ordinary room without that retrofit. The paper's senior author, UC San Diego bioengineer Michael Yip, frames the goal as expanding access to critical surgeries in remote and under-resourced regions. Those are places where installing a fixed console system is impractical.

The Fixed Design Behind da Vinci's Precision

Da Vinci is not a robot that operates on its own. A surgeon sits at a console a few feet from the patient and views a 3D endoscopic image. Hand controls at the console translate one-to-one into motion at the patient-side cart, which holds up to four robotic arms carrying interchangeable EndoWrist instruments. Every motion at the instrument tip is the surgeon's own motion, scaled down and stabilized against hand tremor.

That link between console and cart depends on a mechanical feature called the remote center of motion. It is engineered into each patient-side arm so an instrument pivots around one fixed point at the incision, no matter how the surgeon moves the handles. It is precise because it is fixed mechanically. A general-purpose robot without that fixed pivot has to calculate the same constraint in software instead, continuously, while it moves.

Intuitive Surgical has installed more than 11,000 da Vinci systems worldwide as of early 2026. A da Vinci 5 unit lists for $1.8 million to $2.5 million before instruments and service, according to the American College of Surgeons. That price reflects a single-purpose machine built for one job: teleoperated soft-tissue surgery, cleared procedure by procedure through the FDA. A general-purpose humanoid platform is chasing something da Vinci was never built to be. Giving up that fixed mechanical pivot is the first real cost of the trade.

Why Teleoperation and Haptics Get Harder on a Humanoid Platform

UC San Diego's own team ran into that tradeoff directly. Surgie has no mechanical remote center of motion, so the researchers built a virtual one instead. It tracks visual markers on the robot and runs real-time inverse kinematics, recalculating the pivot point continuously because there is no fixed joint to hold it in place. The Nature paper reports the consequence: the system needed frequent recalibration during surgery and took longer than a comparable da Vinci procedure typically does.

The paper names the gap precisely. Compared with da Vinci, the humanoid setup gave surgeons a less intuitive control interface and more constrained reachability inside the abdomen. Latency between a controller's movement and the robot's response also reduced how responsive the system felt in the surgeon's hands. The authors say they are now testing longer-distance operation, to see how much worse that delay gets over real network conditions rather than inside a lab.

Force feedback compounds the challenge, and it is not unique to humanoid platforms. For most of da Vinci's history, the system gave surgeons no direct sense of how hard an instrument pressed on tissue. Surgeons learned to judge force visually instead, watching tissue deform on the console screen. Intuitive Surgical closed part of that gap itself. The FDA cleared da Vinci 5 with an optional Force Feedback system in March 2024. Intuitive's own preclinical data showed up to 43 percent less tissue force in procedures that used it. A humanoid platform trying to match da Vinci on capability now has to catch up to an incumbent that already has haptics too.

The FDA Pathway a General-Purpose Surgical Robot Would Face

Every da Vinci clearance, and nearly every surgical robot clearance issued so far, has gone through the FDA's fastest track. A systematic review published in npj Digital Medicine found that 44 of 49 FDA-cleared surgical robots, about 90 percent, reached the market through the 510(k) pathway. That pathway requires a manufacturer to show a new device is substantially equivalent to one already cleared. It works well for an incremental da Vinci successor. It works less well for a robot with no real predecessor to compare against.

A precedent already exists for what the alternative looks like. CMR Surgical's Versius system received FDA marketing authorization in October 2024 through the De Novo pathway instead. That is the route the agency uses for a genuinely new type of device with no predicate to point to. CMR describes Versius as compact and modular enough to move between hospital departments, rather than living permanently in one operating room. That is the same portability argument now made for humanoid platforms, minus the humanoid form factor.

A genuinely general-purpose robot raises a harder question than Versius did. FDA clearance covers a specific intended-use statement tied to defined procedures. It does not cover a general claim that a robot can perform surgery. A platform capable of many different operations would likely need a separate clearance for each one, the same way da Vinci does today, procedure by procedure. Liability follows the same pattern for now. The credentialed surgeon at the controls remains accountable for the outcome. The robot's manufacturer does not carry that accountability today. That distinction, and what a hospital has to verify before adopting any teleoperated system, is covered in the guide to how current surgical robots are bought and credentialed.

What Has to Be Proven Before This Reaches a Hospital

None of this has been tried on a human patient. Every procedure in the Nature study was performed on a live animal model under a research protocol. It was not a clinical trial with human consent, and the authors present their own limitations list rather than a launch date. Sterilizing a jointed, motor-driven humanoid body between cases is one item on that list. Da Vinci's arms were engineered from the start to be draped and cleaned for repeat clinical use. A humanoid platform borrowed from other tasks was not built with that requirement in mind.

Cost is the other place the comparison needs care. A base Unitree G1 humanoid retails from about $13,500 in hardware, next to $1.8 million to $2.5 million for a da Vinci 5 unit. That is not yet a fair comparison. Da Vinci's price includes the software, instruments, service contract, and years of regulatory validation that a research robot has not gone through. None of those costs are optional for a system a hospital could actually run.

For now, the realistic reading is narrower than the headline claim. A commercial humanoid robot has proven it can be teleoperated through a real laparoscopic procedure on a living body, which nobody had shown before. Whether it can do that reliably, safely, and cheaply enough to reach a clinic that cannot afford a da Vinci system is still an open research question. It is not yet a product on a price list.

• Human clinical trials, beyond the live-animal feasibility work published so far.
• Latency low enough to hold up over an actual remote or cross-border network connection.
• A sterilization protocol built for a jointed, motor-driven robot body.
• A procedure-by-procedure FDA clearance, or a new device classification built for general-purpose platforms.
• A validated total system price that covers software, service, and instruments, beyond the base robot hardware cost alone.