Robotic-Assisted Retina Surgery: What It Is and Where It Stands Today
Robot-Assisted Retina Surgery at a Glance
Robot-assisted retina surgery is ordinary retina surgery with one change. The surgeon does not hold the tiny instrument by hand. A robot arm holds it, and the surgeon steers that arm with a joystick and a foot pedal. The robot translates the surgeon's large-scale hand movements into precise micro-movements at the instrument tip, and removes the tool movement that comes from hand tremor1.
The surgeon makes every decision and drives every motion. Think of it as a very steady tool holder, not a replacement for your doctor.
The retina is the thin lining at the back of the eye that senses light, and it does not grow back once it is damaged. That makes small slips matter. Every human hand shakes a little, and during retinal surgery that shake is about 100 micrometres, close to the width of a fine hair1.
Skilled surgeons work around this every day, which is why hand surgery works well. A robot arm holds a tool steadier still, and the hope is that this makes delicate steps safer.
Right now this is an early technology, not a routine choice. No eye-surgery robot holds FDA approval in the United States1, so in this country patients meet these systems inside research studies rather than in routine care. In Europe, one system called Preceyes carries a CE mark and has been used in clinical trials1.
If you need retina surgery this year, you will almost certainly have skilled manual surgery, and that is a good outcome rather than a second-best one.
What Robot-Assisted Retina Surgery Actually Is
The systems used in eye surgery so far are what engineers call telemanipulators. The surgeon sits at the microscope and moves a controller, and the robot arm reproduces those movements at a much smaller scale, filtering out hand tremor as it goes1.
A second design, called a cooperative system, works differently. The surgeon holds the instrument and the robot holds it too, damping the shake. That design is still in the laboratory for eye work.
Most people know robotic surgery from general or urology operations, which often use a large system called the da Vinci. That system works to roughly 1 millimetre of precision, which is far too coarse for work on the retina1.
Eye robots were built on a finer scale, working inside a space smaller than a grape through openings under half a millimetre wide.
Several systems exist at very different stages. Only a few have ever been used on a living human eye, and the rest are in laboratories or animal studies.
| System | Who develops it | Where it stands |
|---|---|---|
| Preceyes Surgical System | Preceyes BV, now part of Carl Zeiss Meditec | CE marked in Europe in 2019; used in human trials |
| Luca | AcuSurgical, France | Early human vitrectomy cases in Europe |
| OQrimo | Riverfield, Japan | Endoscope-holding robot, approved in Japan in 2023 |
| Steady-Hand Eye Robot | Johns Hopkins University | Laboratory stage, no human eye surgery yet |
| IRISS | University of California, Los Angeles | Laboratory stage, tested in animal eyes |
Which system a centre has, if any, depends on what study it runs.
Eye robots are designed on the assumption that something will go wrong at some point. Boundaries are set in software so the instrument cannot travel outside the zone the surgeon has defined, the arm retracts automatically if the patient moves unexpectedly, and a foot pedal lets the surgeon release the instrument at once2.
The surgeon also stays at the table throughout, ready to take the instrument out by hand and finish in the usual way.
How a Robot-Assisted Retina Operation Works Step by Step
Almost every robot-assisted retina case begins as a normal vitrectomy. In a vitrectomy, the surgeon removes the clear gel that fills the eye (clinical: the vitreous) through very small openings and replaces it with saline, a gas bubble or oil, usually as a day case taking one to several hours3.
This part is done by hand, exactly as it would be without a robot. The robot is brought in only for the fine step at the end.
The surgeon stays at the operating microscope and looks into your eye as usual. The surgeon works from a console with a joystick-style controller that translates hand movements to the robot arm, and the system passes through the same standard 23, 25 and 27 gauge openings used in hand surgery4.
Because the openings are the same, nothing about your incisions changes.
The first robot task tested in people was membrane peeling. A sheet of scar-like tissue can grow on the macula, the small central patch of retina you read with, and pull it out of shape. In the first human trial, the robot was used to lift and peel that membrane off the retinal surface5.
Lifting the first edge is the hardest moment of that operation, because the tissue underneath is a fraction of a millimetre thick. A steadier tool is meant to help there.
The second task is harder: putting a drug into the paper-thin space under the retina. In an Oxford trial in people with a bleed under the centre of the retina, the robot advanced a fine tube through the retina and held it still while up to 100 microlitres of a clot-dissolving drug was injected by foot pedal6.
Holding a needle motionless in that space is near the limit of what a hand can do. This task most interests retina surgeons, because gene and cell treatments are given the same way.
Who Is a Good Candidate for Robot-Assisted Retina Surgery
Only a narrow set of conditions has been studied in people. The published human trials covered scar tissue on the macula and bleeding under the centre of the retina from wet macular degeneration56. A separate early study used a robot to place a needle into a retinal vein in four people1. Detached retinas and diabetic surgery remain untested.
If your problem falls outside that short list, robot assistance is not on the table today. That does not limit your treatment, because the standard operations for those problems are well established.
Manual retina surgery already works well, and that is the honest reason robots are not spreading fast. After vitrectomy for a macular membrane, roughly 80 of every 100 patients gain at least 2 lines of vision on the eye chart7.
A new tool has to beat that record, not just match it. So far the trials show similar results rather than better ones, a reasonable place for a young technology to start.
One system is available in Europe under a CE mark, an endoscope-holding robot was approved in Japan in 2023, and a French system has treated a small number of patients in early European studies4.
If you want to be involved, the realistic route is a clinical trial at a large academic eye centre. Your retina specialist can tell you whether a study is open near you.
Who Should Wait or Choose a Different Approach
Some retina problems are time-sensitive, and delay costs sight. A detachment involving the centre of vision, or a bleed smothering the macula, needs treatment on the surgeon's normal timetable. Standard vitrectomy is the established treatment for retinal detachment, macular holes and bleeding inside the eye3.
Waiting weeks for access to a robot in that situation would trade a real benefit for an untested one. Have the operation that is available now, and you keep the outcome the evidence supports.
Robot assistance has not been studied in children, in eyes heavily scarred by long-standing diabetes, in urgent detachment repair, or in eyes with unusual anatomy after injury.
Being outside the evidence is not the same as being in danger. It means the honest answer to whether the robot helps you is not yet known, so the standard operation is the sensible choice.
Some reasons are about the operating room rather than your eye. Robot cases take longer, need extra floor space and an extra foot pedal, and require the whole team to be trained and credentialed, and standard training programmes for this do not yet exist14.
If your surgeon says the robot is not right for your case, that is usually a workflow judgment rather than a comment on your eye. It is fair to ask which of the two it is.
What Happens at the Consultation Before Surgery
The consultation looks like any retina surgery consultation. You have your vision measured, drops to widen the pupil, an examination of the retina and a scan of the macula. Vitrectomy is offered when reduced vision, distortion or double vision bothers you enough to justify the risks7.
Bring a note of what you can no longer do comfortably, such as reading a menu or recognising faces. That is the information that decides whether surgery is worth it.
If a robot is genuinely on offer, the consent conversation has an extra layer. You should be told plainly that the robot part is investigational, what happens if the team switches back to hand instruments, and who funds the study.
You are free to say no to the robot and yes to the surgery. Declining the research part never changes the standard care you are given.
These systems are costly to buy and maintain, and because robot-assisted cases are reimbursed at the same rate as manual ones, they are not yet cost-effective for hospitals4.
In a clinical trial, the study usually covers the robot part. Ask what you would be billed for and get the answer in writing before surgery.
Recovering After Robot-Assisted Retina Surgery
Recovery follows the vitrectomy, not the robot. You wear an eye patch for several days, and a scratchy or gritty feeling in the eye is common and settles with time and the drops you are given3.
Vision is blurry at first and clears gradually over weeks. Plan for someone to drive you home and help for a day or two.
Some retina operations end with a gas bubble in the eye to hold the retina in place, though a straightforward membrane peel often does not need one. When a gas bubble is used, you are asked to hold your head in a set position for a set time, and to avoid air travel and high altitude until the bubble has gone3.
Ask your surgeon before booking any flight, because the bubble can take weeks to clear. Following the positioning instructions closely is the most useful thing you can do.
Timing depends on your job, your other eye and whether a bubble was used, so your surgeon sets the dates. As a general shape, desk work often resumes within a week or two, while heavy lifting and swimming wait longer.
Do not drive until your surgeon agrees you meet the vision standard where you live. Asking at your first follow-up visit gets you a clear date rather than a guess.
Risks, Complications and a Realistic Outlook
Nearly all the risk of a robot-assisted case is the risk of the operation underneath it. The risks of vitrectomy include infection, bleeding, a torn or detached retina, poor vision and raised pressure inside the eye, and a cataract can also form in the operated eye, which is especially likely in people over 503.
Most of these have well-worn treatments, and cataract surgery afterwards is routine. Knowing the list in advance makes follow-up visits less anxious.
In one series of 216 eyes that had 25 gauge vitrectomy for a macular membrane, retinal detachment occurred afterwards in 2 eyes, close to 1 of every 1008. In a 20 year review of 111,876 vitrectomies at one hospital, infection inside the eye (clinical: endophthalmitis) followed about 1 of every 10,000 small-gauge operations9.
Both are recognised complications with established treatments, which is why the same-day warning signs below matter. Knowing what to watch for turns a rare event into a manageable one.
The robot brings its own unknowns: longer time under anaesthesia, a machine that could fault mid-case, and a team early on its learning curve. In the first human trial the robot-assisted peel took a median of 4 minutes 5 seconds against 1 minute 20 seconds by hand5.
Those extra minutes are short within a whole operation, and the surgeon can revert to hand instruments at any point. No robot-specific harm has been reported so far, though the trials are far too small to rule out uncommon problems.
For the average person having a macular membrane removed, the outlook comes from the operation rather than the tool. On average about 80 of every 100 patients with a macular membrane or vitreomacular traction improve by at least 2 lines of vision after vitrectomy7, which leaves roughly 20 of every 100 who do not gain that much.
Those are population figures and cannot tell you what your own eye will do, because that depends on how long the macula was distorted. Your surgeon can talk through where you sit in that range.
Warning Signs to Call Your Retina Surgeon About the Same Day
After any retina operation, a few symptoms mean you should call the surgical team the same day rather than wait for your next appointment. Sudden flashing lights, many new floaters at once, a shadow in the side vision and a grey curtain across part of the field of vision are the symptoms of a detached retina10:
- Seeing flashing lights all of a sudden.
- Noticing many new floaters at once, the specks, lines or cobwebs that drift across your view.
- A shadow appearing in your side vision.
- A grey curtain covering part of your field of vision.
- Vision that is getting worse rather than slowly clearing.
Those symptoms are worth acting on quickly. Among people who suddenly develop floaters or flashes from any cause, a pooled review of 17 studies found a retinal tear in about 14 of every 100, with a drop in vision the symptom most strongly linked to a tear, and the authors advise triaging these patients for urgent eye assessment11.
Call about increasing pain or increasing redness in the operated eye as well. Infection is one of the listed risks of vitrectomy3, and your team would far rather check an eye early than late. Most calls of this kind end in reassurance.
Some contact is planned rather than urgent. Patients who have vitreoretinal surgery are routinely examined on the day after surgery, a visit driven by concerns about infection, wound leaks and raised pressure inside the eye12. Your team sets the later visits before you go home, so ask for the dates and write them down.
Keep those appointments even when the eye feels fine, because early changes inside the eye are often silent.
Call the retina surgeon's office or the number on your discharge sheet, not a general helpline, because the team that operated knows what was done. Say which eye, what day you had surgery, what the new symptom is and when it started.
If it is out of hours and your vision is dropping fast or the eye is very painful, use the emergency contact you were given. Being seen and sent home is a good outcome.
What the Research Shows and What Is Still Being Tested
The evidence base is small and honest about itself. The first trial randomly assigned 12 patients, 6 to robot-assisted membrane peeling and 6 to manual surgery, and found surgical success equivalent between the groups5. A second randomised trial of 12 people found robot-assisted injection under the retina was tolerated and safely completed in every case6.
Twenty-four patients in total is a starting point, not an answer. It supports the idea that this is workable and reasonably safe, and shows no benefit for vision yet.
Interest in gene and cell treatment keeps this field moving, because those medicines have to be laid under the retina precisely. In animal-eye laboratory work, a head-mounted robot formed the required fluid pocket under the retina in all 21 injections it attempted4.
That is laboratory work in animal eyes, so it is a signal rather than a result. Human studies would need to repeat it before any of this reaches a clinic.
Four limits are worth holding on to. The trials are tiny, they ran at one or two pioneering centres, they measured safety rather than long-term vision, and they compared a new tool against a mature technique.
A robot completing an operation is news about engineering. Only a larger trial with vision outcomes would be news about your sight.
Common Questions About Robot-Assisted Retina Surgery
No. Every eye robot used in patients so far is directly driven by a surgeon who is scrubbed, at the microscope and in control of every movement. The machine copies the surgeon's hand at a smaller scale and filters out shake. It has no ability to decide what to do. If the system faults or you move, it retracts and the surgeon finishes by hand.
Probably not, unless you live near a large research eye centre in Europe or Japan. In the United States these systems are not FDA-approved, so any use sits inside a study. In Europe one system carries a CE mark that allows its use in care. Your retina specialist can tell you whether a trial is recruiting near you.
There is no evidence that it does, and the trials were not designed to show that. The two randomised trials compared safety and workability, and robot-assisted surgery came out similar to manual surgery rather than better. Vision after macular membrane surgery depends mostly on how long the macula was distorted beforehand.
It is too early to say. The first trial recorded slightly less minor retinal trauma in the robot group, but the difference was not statistically significant, meaning it could easily be chance in a group of 12. Showing a real safety gain would need a trial many times larger than any run so far.
Yes, at least for now. In the first trial the robot-assisted peel took a median of about 4 minutes against about 1 minute 20 seconds by hand, and setting the system up adds more time. Since the delicate step is a small part of a longer operation, the extra time is modest. Teams get faster with experience.
In practice the question rarely arises, because the robot is used in studies where the research budget covers it. Robot-assisted cases are paid at the same rate as manual ones, one reason hospitals have been slow to buy the systems. If a centre offers you robot assistance outside a trial, ask for the billing details in writing first.
More Questions Patients Ask About Eye Surgery Robots
That is the main reason surgeons are interested, though it is not standard practice. Gene and cell treatments have to be placed in a very thin space under the retina, and a robot can hold a needle still there for minutes. One trial has used a robot to inject a drug into that space in people. Whether this improves how well gene therapy works is an open question.
Usually yes. Vitrectomy is normally done with local anaesthesia and sedation, so the eye is numb while you stay awake but relaxed, and general anaesthesia is kept for people who are very anxious or cannot lie still. The robot trials used both: the first membrane-peeling trial was done under general anaesthesia, the later injection trial under local. You should feel pressure rather than pain.
The surgeon carries on by hand. Every case is planned so the robot can be released within seconds using a foot pedal, and the instrument is replaced with the standard hand-held one. This does not put your eye in danger. Ask your team to talk you through their fallback plan if it would settle your mind.
Start with your own retina specialist, who will know which nearby centres run surgical studies. Public trial registers list active studies and their locations, and most robot studies have run at large university eye hospitals in Europe. Be cautious about any clinic advertising robotic eye surgery directly to patients.
- Is the robot part of my operation research, or standard care where you practise?
- What exactly would the robot do, and which parts would you still do by hand?
- How many robot-assisted cases has this team done, and what happened in them?
- What would make you switch to hand instruments during my operation?
- How would my recovery or follow-up differ from standard surgery?
- What will I be charged for, and what does the study cover?
- If I say no to the robot, does anything about my care change?
- Eye (Nature), peer-reviewed review (PMC11885832) (2024). Robotising vitreoretinal surgeries.
- American Academy of Ophthalmology, EyeNet Magazine (2025). When Robots Meet the Retina.
- American Academy of Ophthalmology, EyeSmart patient education (2024). What Is a Vitrectomy?.
- Retina Today (November/December 2025) (2025). Robotics in Retinal Surgery: Recent Advances and Applications.
- Edwards TL, Xue K, Meenink HCM, et al. Nature Biomedical Engineering 2018;2:649-656 (2018). First-in-human study of the safety and viability of intraocular robotic surgery.
- Cehajic-Kapetanovic J, Xue K, Edwards TL, et al. American Journal of Ophthalmology 2022;237:104-113 (2022). First-in-Human Robot-Assisted Subretinal Drug Delivery Under Local Anesthesia: A Randomized Clinical Trial.
- American Academy of Ophthalmology, Preferred Practice Pattern (approved September 2024) (2024). Idiopathic Epiretinal Membrane and Vitreomacular Traction Preferred Practice Pattern.
- Eye (Nature) 2017; retrospective observational case series (PubMed 28524883) (2017). Occurrence rate of retinal detachment after small gauge vitrectomy for idiopathic epiretinal membrane.
- Patel SN, Dave VP, Pathengay A, et al. PLOS ONE 2018;13(1):e0191173 (2018). Incidence, microbiology, and outcomes of endophthalmitis after 111,876 pars plana vitrectomies at a single, tertiary eye care hospital.
- American Academy of Ophthalmology, EyeSmart patient education (2025). Detached Retina.
- Hollands H, Johnson D, Brox AC, et al. JAMA 2009;302(20):2243-2249 (Rational Clinical Examination systematic review) (2009). Acute-onset floaters and flashes: is this patient at risk for retinal detachment?.
- Retina Today (October 2016) (2016). Rethinking Timing of the First Postoperative Visit.