Humanoid robots perform world-first operation on live pigs
UC San Diego surgeons used Unitree G1 humanoid robots to remove gallbladders from live pigs, marking the first preclinical trial of its kind published in Nature.
UC San Diego surgeons used Unitree G1 humanoid robots to remove gallbladders from live pigs, marking the first preclinical trial of its kind published in Nature.
What actually happened in this experiment?
Surgeons at the University of California San Diego remotely controlled humanoid robots to perform two minimally invasive gallbladder removal surgeries on live pigs. The results were published in the journal Nature, making this the first preclinical trial of its kind. The robots were not acting autonomously; human surgeons directed every movement from a control console equipped with a stereo headset display and a foot pedal to engage or disengage their hand motions from the surgical tools.
The team used Unitree G1 robots, a commercially available humanoid platform from Chinese robotics manufacturer Unitree. Because the G1 was not built for surgery, the researchers had to fabricate physical adapters so the robots could grip surgical instruments, and they wrote custom software to translate natural hand movements into precise tool control. The robots were nicknamed "Surgie." The first pig surgery included a human assistant standing beside the robot; the second used two teleoperated robots working together.
Humanoid robots performing surgery on live animals is a proof of concept, not a product, and the gap between those two things is still very wide.
Why does cost and size matter here?
The central argument for this approach is economics and logistics. Intuitive Surgical's da Vinci system, the dominant specialized surgical robot, can cost anywhere from half a million to several million dollars and weighs roughly 1,800 pounds. The Unitree G1 stands five feet tall, weighs 60 pounds, and starts at $13,500, though adding dexterous hands and other necessary upgrades can push the total past $67,000. That is still a fraction of the da Vinci's price.
The size difference matters just as much as the price. Specialized surgical robots occupy significant floor space in operating rooms, which limits their deployment to well-resourced hospitals. A 60-pound humanoid robot is far easier to move and install, which is why the UC San Diego team sees potential use in rural clinics, military field settings, and even space missions. Shanglei Liu, an assistant professor of surgery at UC San Diego School of Medicine, said the system is "a fraction of the cost and it takes a fraction of the space in an operating room" and is "easy to deploy, anywhere from rural areas, to the battlefield, and even to space."
What are the real limitations right now?
The surgeries worked, but they were slow and demanding. The team had to pause multiple times during each procedure to recalibrate the robots or physically reposition them, because the Unitree G1's arm span is only 450 millimeters, far shorter than the 1.6 to 1.8 meter reach of an adult human. That constraint, combined with limited range of motion, forced the surgical team to manage higher cognitive and operational workloads than they would face with existing specialized systems. The total operating time was significantly longer than comparable procedures performed with conventional surgical robots.
Latency is another open problem. Any delay between a surgeon's hand movement and the robot's response becomes critical when the procedure involves remote operation over a network. The current experiment does not appear to have resolved that challenge. The da Vinci system, by contrast, has FDA clearance and a track record across multiple clinical trials. The humanoid robot approach remains firmly in the experimental phase, and a path to regulatory approval for human patients has not been established.
What does this mean for the robotics industry?
This experiment matters less as a near-term medical product and more as a signal about where general-purpose humanoid robots are heading. The fact that a commercially available humanoid platform, with hardware modifications and custom software, can perform a real surgical procedure on a living animal is a meaningful capability milestone. It suggests that the gap between industrial humanoid robots and specialized task robots is narrowing, even if it has not closed.
For robotics companies building humanoid platforms, surgical applications represent a high-value, high-scrutiny proving ground. Success there would validate dexterity, precision, and teleoperation software in ways that warehouse or logistics deployments cannot. For hospitals and health systems, the more immediate question is whether a low-cost, portable teleoperation platform could extend surgical access to underserved areas before a fully autonomous surgical robot ever reaches the market. That is a different value proposition than replacing the da Vinci, and it may be a more achievable one in the near term.
How does the Unitree G1 humanoid robot compare technically to a dedicated surgical robot like the da Vinci?
The da Vinci system is purpose-built for surgery, FDA-cleared, and has been validated across many clinical trials, but it can cost several million dollars and weighs around 1,800 pounds. The Unitree G1 is a general-purpose humanoid that costs under $70,000 fully equipped, weighs 60 pounds, and required custom adapters and software before it could hold surgical tools. Its arm span of 450 millimeters is far shorter than a human's, which limited reach and required frequent repositioning during the pig surgeries.
Could this approach actually make surgery available in places that currently lack access?
That is the core promise, but it is unproven at scale. The UC San Diego team argues that a lightweight, low-cost humanoid robot is far easier to deploy in rural clinics, battlefield settings, or remote locations than a multi-million-dollar specialized surgical system. The preclinical pig trial shows the concept is physically feasible. Whether it can be made fast enough, safe enough, and reliable enough for human patients in those settings is a question that will require many more trials and eventual regulatory review.
What needs to happen before humanoid robots could be used in surgery on human patients?
Several major hurdles remain. The surgical times need to come down significantly, as the pig procedures took much longer than equivalent operations with existing systems. Latency between surgeon input and robot movement must be minimized, especially for any remote or networked scenario. The robots' limited arm span and range of motion need engineering solutions. Most importantly, the system would need to go through formal clinical trials in humans and obtain regulatory clearance from bodies like the FDA, a process that typically takes years and requires extensive safety data.