UC San Diego researchers demonstrated that a $13,500–$67,000 off the shelf humanoid can execute full keyhole (laparoscopic) surgery under remote human control — an early feasibility step that reframes what surgical robots are for and where they can
Surgeons at UCSD have used a teleoperated humanoid robot to perform the first peer-reviewed laparoscopic surgery on a live pig, a milestone the field has not previously cleared on a living animal. The result, published in Nature and corroborated by a UCSD press release, reframes the surgical-robotics story around spatial decoupling: letting a surgeon operate from a distance. The robot is not thinking. The surgeon is just not in the room.
The platform is a Unitree G1 humanoid, priced between roughly $13,500 and $67,000 depending on configuration. An Intuitive Surgical da Vinci system runs $500,000 to more than $2 million, plus dedicated operating-room infrastructure. Study author Dr. Shanglei Liu told Ars Technica the humanoid form factor is meant to be deployable. "It's a fraction of the cost and it takes a fraction of the space in an operating room. So it's easy to deploy, anywhere from rural areas, to the battlefield, and even to space."
Spatial decoupling is the underlying mechanism. Existing surgical robots like the da Vinci keep the surgeon in the same room as the patient, sitting at a console. A teleoperated humanoid can be sent somewhere the surgeon cannot easily reach, with the surgeon controlling every cut and suture from a remote interface. The UCSD team demonstrated that workflow in vivo: a surgeon drove the humanoid through a full laparoscopic soft-tissue procedure on a sedated pig, with the robot executing the motions under continuous human control. The study design is a feasibility test, not a comparison against a da Vinci baseline, and the team is explicit that the work is preclinical.
The project's institutional site describes the system as a research platform for studying how general-purpose humanoids can be adapted to surgical tasks, rather than a finished clinical device. Coverage from Medical Xpress and Ars Technica lines up with the Nature paper and the UCSD release: teleoperation, in-vivo work on live pigs, no human patients, no published timeline for clinical trials.
The available reading is that humanoid robots can now perform surgery on their own. They cannot. Every instrument move in the study came from a human surgeon at a control station. What the platform has actually cleared is a feasibility hurdle the da Vinci class of machines cleared decades ago, on a different body, and the next gate is years of refinement, animal work, and regulatory review before any human use. Liu's deployment pitch is forward-looking, not an operational claim.
The Nature paper is the durable record. The cost data and the deployment rationale live in Ars Technica's reporting. What to watch next: comparative studies against conventional laparoscopic systems, failure modes across more cases, and any IDE application a regulator sees for a humanoid surgical platform.