Robotic-Assisted Knee Replacement in Perth

Robotic assistance has become one of the most talked-about advances in joint replacement, and one of the most misunderstood. The word invites a picture of a machine operating on its own while you are asleep.

That is not what happens. In a robotic-assisted knee replacement your surgeon performs the operation and makes every decision. The robotic system carries out a plan the surgeon built beforehand — from a CT scan of your knee, or mapped in theatre depending on the system — and provides real-time feedback during surgery so the implant is placed exactly where it was planned.

This page explains how that planning works, what the precision is actually for, what the evidence supports and what it does not yet, and how to decide whether it matters for you.

What does the robot actually do?

It holds the plan, and it helps execute it accurately.

Before your operation, Dr Clark builds a three-dimensional model of your knee and decides implant size, position and alignment down to the millimetre. During surgery the robotic arm guides the bone preparation and implant placement against that plan, giving continuous feedback on where the instruments are relative to where they were meant to be.

What it does not do is decide anything. It does not assess your soft tissues, judge how the knee is balancing, or adapt to something unexpected — those are surgical judgements, and they remain with the surgeon throughout. If the joint turns out differently once it is opened, the plan is adjusted then and there.

“Before any surgery takes place, I use the images from your CT scan to complete a virtual pre-operation on a computer. When it comes to performing your surgery, the robotic arm helps guide me in making the bony cuts and implant placement,” says Dr Rhys Clark.

The useful analogy is a GPS. It knows the route, it tells you when you are drifting off it, and it makes the journey more precise. It is not driving.

How your operation is planned before you arrive

Where your operation is planned on the Mako system, you will have an additional CT scan, separate from the X-rays and imaging used to assess how much arthritis is in the knee. This one exists purely for planning. (The VELYS system, which Dr Clark also uses, works the other way round — the knee is mapped in theatre and no planning scan is needed. Both are covered below.)

From it, a three-dimensional model of your knee is built — your bone shape, your alignment, your deformity if you have one. Dr Clark then rehearses the operation on that model: which implant size fits your anatomy, exactly where it should sit, how the alignment should be corrected, how the components will relate to one another through the full arc of movement.

By the time you are in theatre, the significant decisions have already been made and tested. That is the real shift robotic assistance introduces — not that the surgery is done differently, but that it is planned in far more detail, on your specific knee, before an incision exists.

Why an imaging background changes how a knee gets planned

This part is specific to Dr Clark rather than to robotics generally, and it is worth knowing if you are comparing surgeons.

Dr Clark came to orthopaedics through medical imaging. Alongside his fellowships with the Royal Australasian College of Surgeons (FRACS) and the Australian Orthopaedic Association (FAOrthA), he holds a Master of Medical Radiation Science and a Bachelor of Applied Science in Medical Imaging.

Robotic knee replacement is, at its core, an imaging problem before it is a surgical one. The quality of the operation is bounded by the quality of the plan, and the plan is built entirely from a CT dataset — how it is acquired, how it is interpreted, and how confidently its limitations are understood. A surgeon with formal training in the modality is reading that scan with a different depth of background than one who came to it only through surgical practice.

It is not a claim that this produces a better knee. It is a straightforward statement about where the planning expertise comes from.

What does the precision actually buy you?

It is tempting to think a knee replacement is a knee replacement. But small differences in alignment and implant fit change how the joint feels day to day.

A knee that is even slightly out of balance can produce:

• Stiffness, or a joint that never quite settles
• An unnatural feeling through the gait
• Uneven loading across the implant, and faster wear over time

Robotic assistance narrows the margin for error and allows a fit matched to your own bone shape, ligament tension and range of movement rather than a generic template. Surgeons using these systems report more predictable alignment and a more natural-feeling knee.

There is also a point about how much bone comes out. A total knee replacement removes around 5 to 7 mm from the ends of the bones — considerably less than most people expect — and accurate planning is part of keeping that figure where it belongs.

Which robotic system — Mako, ROSA, VELYS and CORI

Patients often arrive having read about one system by name, usually because a hospital or a practice markets that one. It is worth knowing what actually separates them, because the real difference is narrower than the marketing suggests and it comes down mostly to a single question: where the plan comes from.

Systems planned from a scan taken beforehand. Mako (Stryker) builds its plan from a CT scan taken before surgery, and uses a robotic arm with haptic feedback — the arm physically resists being moved outside the planned boundary. ROSA (Zimmer Biomet) can also work from images, though it generates its three-dimensional model from standard X-rays rather than a CT, and can alternatively be run without any prior imaging.

Systems that map the knee in theatre. VELYS (DePuy Synthes) and CORI (Smith+Nephew, the successor to the Navio system) are imageless. Instead of planning from a scan, the surgeon maps the anatomy with a probe once the knee is open, and the model is built from those points. That avoids an extra scan, and the radiation dose that comes with it.

Both approaches are in routine use and both are supported by evidence. The image-based route front-loads the planning, so more is settled before theatre; the imageless route removes a scan and an appointment. Comparative studies between the major systems have generally found the differences in accuracy between them smaller than the difference between robotic assistance and conventional instruments.

The systems Dr Clark uses. Dr Clark uses Mako and VELYS — one from each of the two groups above.

Mako is the CT-planned system, and it is the workflow described earlier on this page: an additional scan beforehand, a three-dimensional model of your knee, and the operation rehearsed as a virtual pre-operation before an incision exists. The arm’s haptic feedback then physically resists being moved outside that plan.

VELYS is imageless. There is no planning scan; the knee is mapped with a probe once it is open, and the model is built from those points during the operation itself.

Which system is used for your knee depends on the hospital and the theatre your surgery is performed in. It is worth asking about, and it is worth knowing in advance, because it determines whether you will need a planning CT scan beforehand.

Robotic assistance, computer navigation and conventional instruments

These three get conflated constantly, including in advertising. They are genuinely different things, and the distinction is useful if you are comparing what surgeons offer.

Conventional instruments. Bone cuts are guided by mechanical jigs, positioned using alignment rods referenced off the anatomy, and adjusted by the surgeon’s assessment in theatre. This is how the great majority of knee replacements have been done, and it produces good results in experienced hands. Its weakness is that a small error in where a guide sits becomes a small error in the cut.

Computer navigation. Trackers are fixed to the bone and a camera in theatre reports, in real time, exactly where the instruments are relative to the alignment target. It measures and it tells you when you are drifting, but it does not constrain the saw — execution remains entirely manual. Navigation has been in use since the early 2000s and reduced alignment outliers when it arrived.

Robotic assistance. Adds execution to the guidance. The plan is held by the system, and the arm or handpiece actively helps keep bone preparation within it — on some systems by physically resisting movement outside the planned boundary. Every decision, and the operation itself, still belongs to the surgeon.

The honest summary is that each step along that line has narrowed the margin for error, and none of them replaces the judgement of the person operating.

What the evidence shows, and what it does not

Worth being straight about, because the marketing around robotics tends to run ahead of the data.

Well established: robotic assistance improves the accuracy of implant positioning and limb alignment. Surgeons hit their intended targets more consistently, with fewer outliers — cases that land well outside the intended range.

Not yet established: whether that accuracy translates into implants lasting longer. Answering that properly requires 15 to 20 years of follow-up, and the technology has not been in widespread use long enough to have produced it. The Australian Orthopaedic Association National Joint Replacement Registry records every joint replacement performed in Australia and is monitoring robotic-assisted procedures specifically for this reason. The early signals are encouraging. They are also early.

Nor is there good evidence that robotic assistance makes recovery dramatically faster or less painful. Recovery follows much the same path either way.

The honest summary: robotics improves the thing it is designed to improve, and the downstream benefits are plausible and being measured rather than proven. Any surgeon telling you robotic assistance guarantees a better outcome is ahead of the evidence.

Is it right for you?

If you are a suitable candidate for a conventional knee replacement, you are generally also a candidate for a robotic-assisted one. It is not reserved for complex cases.

It can be particularly useful where the anatomy is harder to work with — significant bow-leg or knock-knee deformity, unusual bone shape, or a previous injury or fracture that has changed the alignment of the leg. These are the situations where a generic approach has the most room to drift and detailed planning has the most to offer.

The question worth bringing to your consultation is not “should I ask for the robot” but “what approach do you recommend for my knee, and why?” A good surgeon will explain their reasoning either way, and will be honest about what to expect. The experience and judgement of the person operating still matters most.

Total or partial — does it apply to both?

Both. Dr Clark uses robotic assistance for total knee replacement and partial knee replacement alike.

If anything, accuracy matters more in a partial replacement. Only the damaged compartment is resurfaced, so the new component has to sit correctly alongside the parts of your own joint being preserved — there is less room for error when most of the knee is staying put.

Which operation suits you depends on how many of the knee’s three compartments are affected. A partial suits damage confined to one; a total is generally used where two or more are involved.

What does it cost in Perth?

Robotic assistance is part of how Dr Clark performs the operation, not an optional extra billed on top.

Dr Clark works on a no-gap basis for privately insured knee replacement patients, matching his fees to what the insurer pays. Hospital and prosthesis costs depend on your pathway and your fund. The cost of a knee replacement page sets out the public, private and self-funded routes in full, including typical self-funded figures.

How often Dr Clark does this operation

Dr Clark performed 401 hip and knee replacements in 2025, and more than 900 since the start of 2024. Around 60% of that work is knee replacement, and both total and partial knee replacements are performed with robotic assistance.

How often a surgeon performs an operation is a fair thing to ask, and a fair thing to expect a straight answer to.

Where it is performed

Dr Clark consults at St John of God Medical Clinic, Murdoch and at consulting rooms in Mandurah, serving Perth’s southern suburbs and the Peel region.

Surgery is performed at St John of God Murdoch, Sir Charles Gairdner Hospital, Osborne Park Hospital and Perth Children’s Hospital.

What recovery looks like

Much the same as a conventional knee replacement. Most patients bear weight within hours of surgery, stay two to four nights, use a walking aid for the first few weeks, and are reviewed at around two weeks to check the wound and adjust pain management.

Knee Replacement Recovery sets out the whole timeline week by week, including when driving becomes possible and what to phone the rooms about.

Next steps

If you are weighing up knee replacement and want to understand what approach suits your knee, book a consultation or phone the rooms on (08) 6332 6365. If you are earlier in the process and still deciding whether surgery is the right step at all, Understanding Knee Arthritis is the better place to start.