Psst… We received a tip…

Your surgeon is sitting at a robot before they ever get near you.

A tiny virtual instrument moves through a digital "patient", practising that crucial tricky move they'll have to do for you tomorrow. And something is watching them back, software tracking where they are looking, how they move, when they hesitate, and any mistakes.

It tells your surgeon something crucial, that when they perform the maneuver, they actually need to be attentive to what's happening slightly to the left, watching for a nearby complication. And when that happens in your surgery, the surgeon saw it and was ready to respond immediately.

This week’s patent application goes beyond human surgical mentors. Intuitive Surgical Operations, the company behind the da Vinci surgical robot, imagines an AI mentor for robotic surgery training that catches what human mentors may miss.

This week’s published patent application was filed by Intuitive Surgical Operations. The company behind the da Vinci, a robotic surgery system used in over 3 million procedures a year, and this week's patent imagines an AI mentor for training the surgeons who use it.

There are really two machines hiding inside this patent.

The first is the simulator.

Intuitive’s system starts by collecting data from medical procedures. It can capture how instruments moved, how fast they moved, how much force or torque was applied, where expert surgeons looked, and how a trainee performed over time.

Then it builds a simulated procedure. Think of a digital body, digital instruments, and a practice version of the operation. The trainee sits at an actual da Vinci surgeon console, using the same hand controls and foot pedals they'd use in a real operation, while the system watches what they do inside a simulated patient.

This is where the mentor part kicks in. The patent says the AI mentor can be modeled on expert surgeon data. So instead of just saying “good job” or “try again,” the system can show what an expert-style move looks like inside the same simulated space. The trainee does the task. The AI compares the trainee’s movements, timing, attention, and performance against the expert pattern.

That eye-tracking piece matters.

A human mentor can watch your hands, or the screen, and they can tell when you mess up. But they cannot always multitask or know exactly where you were looking, how long you looked there, or whether you missed the thing slightly off to the side.

This system can turn expert gaze into a heatmap. It can show where the expert looked during the procedure, then compare that with where the trainee looked. This is important, as surgery is also about noticing small details before the wrong thing happens.

Also, if the trainee keeps struggling with one step, the system can generate a training path around that weak spot. A custom digital anatomy with a custom task and repeat loop until the skill is fine-tuned.

The second machine is the video tutor.

The patent also describes a system where a clinician can ask questions about surgical videos or simulated procedures. The software pulls from a knowledge base of medical papers, procedure videos, robotic data, and other resources. Then a generative AI model answers the question with citations.

Intuitive is clearly aware of the obvious problem here that you do not want a confident chatbot making up medical advice. So the patent describes guardrails that check whether the AI’s answer is grounded in the underlying sources. If the answer looks unsupported or hallucinated, the system can suppress it instead of showing it.

There is also a clever timing trick. If a trainee pauses a surgery video at minute eight and asks, “What should I be watching for here?”, the system can stop the AI from looking at minute twelve. That means the answer is based only on what the trainee has seen so far, not on a future complication the model secretly peeked at.

Becoming a surgeon takes 13 to 18 years. Residency alone runs five to seven years, with robotic surgery adding another layer of specialist training on top. And every hour a junior surgeon spends learning in the operating room is an hour a senior surgeon spends supervising rather than operating.

That tradeoff matters because the US is already running short. The AAMC projects a shortage of between 19,800 and 29,000 surgeons by 2030, and the bottleneck is training capacity. 

An AI mentor that can run practice sessions, track weak spots, and generate custom drills hands back time to the surgeons doing the teaching.

The market has pieces of this patent, such as VR surgery practice, robotic simulators, video review, AI analytics, and remote mentoring. But Intuitive’s patent tries to bundle the whole thing into one loop, of expert data in, trainee data out, weak spot found, custom practice generated, and questions answered with sources.

Osso VR is one of the obvious names. It is a virtual-reality surgical training company whose platform is used by more than 100,000 healthcare professionals, and its pitch is hands-on procedural practice with assessment. Osso is trying to make surgery practice move aware from lecture-style, and towards doing reps in a safe digital room.

FundamentalVR is in the same “flight simulator for surgery” lane. Its Fundamental Surgery platform uses VR and haptics, meaning trainees can feel some simulated resistance while they practice. The company also talks about tracking training and assessment through its learning platform.

Surgical Science is probably the closest on robotic surgery simulation. Its RobotiX Mentor simulator is built for surgeons to practice the motor and thinking skills needed for robotic surgery, from basic robotic tasks to full virtual procedure modules. Intuitive’s patent goes further by adding the AI mentor layer involving expert gaze, expert movement patterns, personalized feedback, and custom next lessons.

Then there is the surgical video crowd. Medtronic’s Touch Surgery ecosystem is an AI-powered surgical video and analytics platform that can capture videos and turn them into progress benchmarks, procedure metrics, anatomy identification, and instrument time-in-view. Proximie, Caresyntax, and Theator are also chasing this bigger idea that the operating room should become searchable, measurable, and teachable through video and data.

The real market is robotic surgery itself. Intuitive says its systems were used in more than 3.2 million procedures in 2025, up 19% from 2024, with more than 12,100 Intuitive systems installed at customer sites by year end. That is a lot of surgeons, hospitals, trainees, and nervous patients moving through one robotic ecosystem.

And the robot wave is still early. NHS England projects robotic-assisted operations will rise from 70,000 in 2023/24 to 500,000 per year by 2035, with robot assistance used in 9 out of 10 keyhole surgeries within the next decade. More robots means more surgeons who need to learn the robot, fast.

That is where simulation gets interesting. Grand View Research estimates the surgical simulation market was worth US$530.7 million in 2025 and could reach US$1.8 billion by 2033, growing at 16.6% a year. The wider medical simulation market is even bigger, estimated at US$3.5 billion in 2025 and projected to hit US$7.23 billion by 2030.

Intuitive already spent more than US$1 billion on R&D in 2025, and its annual report says its 2026 priorities include platform growth and more digital features.

Surgical training is becoming a software market. Medtronic bought Digital Surgery, a company building surgical AI, data, analytics, and digital education tools, to strengthen its robotic and laparoscopic surgery strategy. Surgical Science bought Simbionix, a medical simulation business from 3D Systems, for US$305 million.

VC money has been sniffing around the same idea. Osso VR, a virtual-reality surgical training and assessment platform, raised US$66 million in Series C funding led by Oak HC/FT. FundamentalVR, a medical simulation company that uses VR and haptics so trainees can feel some simulated resistance, raised US$20 million in Series B funding led by EQT Life Sciences.

Then there is the surgical video crowd. Caresyntax, a vendor-neutral platform that turns operating room data into insights, raised US$180 million in 2024, including US$80 million of equity and up to US$100 million of growth debt, partly to fuel M&A. Proximie, a connected surgery platform that lets surgeons virtually join operating rooms, raised US$80 million in Series C funding. Theator, an AI platform that analyzes surgical videos, raised a US$39.5 million Series A. Everyone is circling the same prize of turning surgery into searchable, measurable, teachable data.

The robotics deals are even bigger. Johnson & Johnson bought Auris Health, a surgical robotics company focused on endoscopic procedures, for about US$3.4 billion upfront, with up to US$2.35 billion more tied to milestones. Stryker bought MAKO Surgical, a robotic arm-assisted orthopedic surgery company, for about US$1.65 billion. Zimmer Biomet bought Monogram Technologies, an AI-navigated orthopedic robotics company, in a deal announced at about US$177 million in upfront equity value.

The first risk is trust.

If an AI mentor teaches a surgeon the wrong lesson, who owns that mistake? The trainee who followed it? The hospital that used it? Intuitive, because it built the system? The patent describes AI answers, citations, guardrails, and feedback based on surgical data. But a cited answer can still be incomplete, outdated, or too confident for the moment.

The second risk is false confidence.

A simulator can measure gaze, timing, movement, hesitation, and performance. Real surgery is complicated in ways that cannot be digitally replicated, such as tissue bleeds, strange anatomy, stress, tired surgical teams… the list goes on. The danger is that a beautiful score in simulation starts to feel like proof of real-world readiness before it has earned that trust, resulting in a false confidence in surgeons.

Or, is this a signal that we are moving towards more automated surgeries?

Your next surgeon will still need human mentors, real practice, and hard-earned judgment. Intuitive is a nice addition.

Intuitive is adding a new kind of training layer. The robot can measure every glance, pause, wobble, and weak spot, then turn those tiny patterns into a better lesson. 

So, read this one for yourself. The patent application is US 2026/0064916 A1, filed by Intuitive Surgical Operations.