Skip to Main Content

Can a robot operate on its own? Yes, says a team from Children’s National Medical Center and Johns Hopkins University. Their Soft Tissue Autonomous Robot stitched together pieces of pig bowel without being guided by a human surgeon.

The news hasn’t sent surgeons scurrying to find new jobs. They see the promise of this technology, but also envision many hurdles for it.


Monique A. Spillman: Surgery will always need the human touch
Jennifer Tseng: Robots could advance surgeons, but won’t replace them
John C. Alverdy: New system isn’t better than a human surgeon — yet
Sherry M. Wren: Robotics and augmented reality will improve surgery

Monique A. Spillman: The key word in the title of the report on autonomous robotic surgery is “supervised.” The researchers drew the conclusion that the robot performed better than human surgeons. But if that’s true, it was only because humans corrected the planned robotic suture path almost half of the time.

The concept of “Do no harm” should be at the centerpiece of developing and deploying robots in surgery. If a robot does something better than I can, I would want it to be part of an operation. At least for now, though, robots can’t do this. Nor can they design an overall plan or make the kinds of minute-to-minute adjustments that can mean the difference between life and death.


I’m a strong believer in the human touch in medicine. In my practice as a gynecologic oncologist, I often do robot-assisted surgery. But sometimes I need to put my hands on the patient because the sense of touch is critical to finding the right path and protecting essential structures. Robots are able to “see” very well, but lack a sense of touch. Until they get better haptic feedback, the human role in surgery will continue to be essential.

In the 1950s, writer Isaac Asimov put forward Three Laws of Robotics in his short story collection “I, Robot.” Two years ago, a colleague and I applied those “laws” to training in robotic surgery. The first two — A robot may not injure a human being, or, through inaction, allow a human being to come to harm and A robot must obey the orders given it by human beings except where such orders would conflict with the First Law — definitely apply to autonomous robots in surgery.

Autonomous machines are developing quickly. A Tesla automobile owner told me that in a few years, the car may be able to drive from Los Angeles to New York all by itself. With sufficient research and development, autonomous robots could represent the next generation of robotic surgery. If they do, truly autonomous robotic surgery will raise new ethical issues in informed consent issues for the robot as the primary surgeon. It will also require new definitions for surgical competence, credentialing, and privileges to perform surgery.

Monique A. Spillman, MD, is clinical associate professor and attending gynecologic oncologist at Baylor University Medical Center.

Jennifer Tseng: Autonomous robotic surgery is an exciting concept. But it’s a long way from that concept to clinical use.

Although the Children’s National Health System team calls the robot it developed “autonomous,” it really isn’t, at least not in the commonly understood sense of the word. It does a complex but repetitive task that human beings expertly programmed it to do.

My colleague Dr. Jim Moser has called the current generation of surgical robots “million dollar needle drivers.” When we do a robotic operation like a Whipple procedure, the surgeon sits at a console and moves his or her hands. That moves the instruments attached to the robot’s arms. The Smart Tissue Autonomous Robot (STAR) takes this a step further. Humans designed and programmed STAR, placed the robotic arm and instrument in place, set up the bowel in the correct way, turned on the robot, and “told” it to sew. Only then did the robot stitch together sections of bowel that had been surgically separated. Make no mistake — this is a technically demanding task. It requires making tiny sutures spaced close enough together to prevent liquid from leaking out of the bowel, but far enough apart to let the bowel heal. The elasticity of bowel tissue makes this more complicated than surgery on bones or other rigid tissue.

The new robot did this well, although it took longer than standard surgical bowel repair.

The most important part of surgery is decision-making. It’s not enough to make a stitch. You have to know where to sew and what to sew, as well as how or why to sew. Even the most perfectly programmed robot can’t do that.

If autonomous robots are perfected for surgery, I can absolutely see them assisting surgeons with technically demanding repetitive tasks like suturing, particularly in minimally invasive settings. To be futuristic, it would be wonderful to develop miniature robots that could work with surgeon guidance in anatomic and physical locations surgeons have difficulty getting to.

Robots will never be substitutes for surgeons, but could be useful extensions of them. Surgeons will have to decide why to operate, and what to do; robots can help with the “how.” This would extend the ability of surgeon to help patients.

Jennifer F. Tseng, MD, is chief of the Division of Surgical Oncology at Beth Israel Deaconess Medical Center and associate professor of surgery at Harvard Medical School.

John C. Alverdy: The idea of an autonomous robot for surgery is interesting, enticing, high-tech, and sexy. While the foundational science behind this report is sound, the outcome data presented in it do little to convince today’s surgeons that we are about to be replaced by robots.

The robot took six times longer than standard open surgery, and made more mistakes. It also shrank the opening in the bowel through which food and liquid pass significantly more than standard surgery — by one-quarter to one-third.

Perfecting an autonomous robot to reconnect two ends of the intestine after a segment is removed, or connecting tiny blood vessels together when replacing or transferring tissues, is an exciting idea to prevent known complications, such as leakage and narrowing or blockage of the vessel. While meticulous technique is always the goal, even if surgeons must relinquish the sewing to a robot, it is important to keep in mind that many of the major complications following such procedures are caused by factors other than surgical technique. These include the patient’s underlying biology and clinical circumstances, such as malnutrition and the size of a tumor. Even the patient’s microbiome, as we have shown, can affect healing of these connections.

That said, this experimental robotic system represents a terrific model for how bioengineers, physicians, surgeons, and others should explore new ways to improve patient outcomes. While this report does little to convince us that it is better than what we do now, it represents the potential to be better.

A robotic system could possibly measure things that surgeons can’t see with their eyes or feel with their fingers. It could also work in fixed spaces and with delicate structures that require precisely made connections. I could see a next-generation robotic system being used by plastic surgeons to reconnect tiny blood vessels, such as is needed in breast reconstruction after cancer surgery. It could also play a role in brain surgery, where one millimeter of movement can have a negative effect.

It remains to be seen if anyone will be able to program a robot to play the violin like Itzhak Perlman, with the same vibrato, phrasing, and dynamics typical of this master musician. In the same sense, it would be a major challenge for a robot to replace a master surgeon with years of experience who sees, feels, and understands tissue textures and dimensions that are difficult to teach and transfer, let alone mechanically recapitulate. That said, this report excites the senses and opens the door for this very possibility.

John C. Alverdy, MD, is professor of surgery and executive vice-chair of the Department of Surgery at the University of Chicago.

Sherry M. Wren: I find it fascinating to be writing seriously about a technology that sounds like it is out of “Star Trek.”

The robot described in Science Translational Medicine isn’t the first autonomous one developed for surgery. The Mako system has been used for several years in orthopedic surgery. It autonomously drills a precise hole in a bone so it can receive an artificial hip or knee. This system is more like an industrial robot that works on a fixed surface. Dealing with bowel tissue or blood vessels represents a higher level of difficulty.

My concern about using a robot in gastrointestinal surgery is that it would disrupt the flow of the operation. The surgeon would have to stop the operation, deploy the robot to do its thing, un-deploy the robot, and then move on to the next stage of the procedure. That could increase the time the patient is under anesthesia.

I can’t see any intestinal surgeon using an autonomous robot for something unless the robot improves the outcome. But robots could play important roles in other operations where precisely joining two pieces of tissue is difficult but essential. Pancreatic cancer is an example. The surgeon needs to connect the pancreatic duct to the bowel. Leaks occur in about 10 percent of operations when the duct is very small. A robot that can make the connection and reduce leaks would be a true breakthrough, one worth disrupting the flow of an operation.

I believe that robotics and augmented reality, such as fluorescent imaging that lets a surgeon better visualize a tumor or blood flow, will make surgery more accurate and safer. Small advances such as the Soft Tissue Autonomous Robot are a step in that direction.

Sherry M. Wren, MD, is professor of surgery and director of global surgery at Stanford University (@sherrywren).