Spine Surgery Robot: Mazor, Excelsius and ABB

A spine surgery robot has a narrow job. It holds a drill guide on the path the surgeon planned on a CT scan, so each screw enters the vertebra where it was meant to go. This post follows that job from freehand pedicle screws checked with X-ray, through the first Mazor SpineAssist in 2004, to the Mazor X, ExcelsiusGPS and ROSA Spine platforms in use now. It also covers recent news from Ontario and Seoul. ABB robots come in at the end, in the factories that make spinal implants; none of them works in the operating room.

An X-ray of a spine held straight by two rods and a column of pedicle screws. Photo: Nevit Dilmen / Wikimedia Commons (CC BY-SA 3.0)

How surgeons placed spinal screws by feel and fluoroscopy

Spinal fusion locks two or more vertebrae together so they heal as one bone. From the 1950s, Houston surgeon Paul Harrington hooked steel rods onto the spine to straighten scoliosis curves, and his rods stayed a standard treatment for decades. Later, surgeons such as Raymond Roy-Camille in France fixed plates and rods with screws driven through the pedicles, the two narrow columns of bone at the back of each vertebra. A pedicle screw grips much better than a hook. The catch is that the pedicle is small, and the spinal cord and nerve roots sit a few millimetres away.

A standing X-ray of a teenager's spine with a scoliosis curve before surgery, University of Utah Hospital, 2009. Photo: University of Utah Hospital Radiology via Wikimedia Commons (public domain)

For years surgeons placed those screws freehand. They found the entry point from bony landmarks, felt the path with a probe, and checked their work with a mobile C-arm that took fluoroscopy images during the operation. Skilled surgeons get good results this way, but it gets much harder with a twisted scoliosis spine, a revision case or a small child. Every check shot adds radiation for the patient and for the staff standing at the table, so spine teams wanted a better view of where the drill was going.

A mobile C-arm fluoroscopy unit and an endoscopy tower in an operating room, 2010. Photo: Doctorqmd / Wikimedia Commons (CC BY-SA 3.0)

Navigation came first. Cameras track markers on the instruments and show their position on CT images, and the surgeon watches the tool move inside the bone on a screen. A robot was the next step. The SpineAssist, made by Mazor Robotics of Caesarea, Israel, was cleared by the FDA in 2004 and is usually cited as the first spine robot. It was a small device mounted on the patient that pointed a guide along the planned trajectory. Its successor, Renaissance, arrived in 2011 with a smaller body and faster software.

A surgeon operating on the spine with the Mazor Renaissance robotic system, 2012. Photo: Ap2296 / Wikimedia Commons (CC BY-SA 3.0)

Mazor X, ExcelsiusGPS and ROSA in spine rooms now

The current generation stands on a cart or clamps to the bed rail and works with a camera. Mazor X, released in 2016, uses its camera to check its own position and registers each vertebra separately. Medtronic bought Mazor in 2018 and pairs the robot with its O-arm imaging. ROSA Spine won FDA clearance in 2016. Globus Medical's ExcelsiusGPS was cleared in August 2017; it combines a rigid arm with navigation and places screws without guidewires, and in September 2026 Globus announced a CE mark for its Excelsius3D imaging system. Reviews report better accuracy and less radiation, while patient outcomes after a year or two look much the same as with freehand surgery.

Medtronic's O-arm mobile imaging system, the first in Russia, at the Federal Neurosurgery Center in Tyumen, 2013. Photo: Dmitry Koshelev / Wikimedia Commons (CC BY-SA 3.0)

Two recent stories show the robots moving into new kinds of cases. Children's Hospital at London Health Sciences Centre became the first hospital in Ontario to use the Mazor platform for a paediatric scoliosis correction, on a 14-year-old girl. Her surgeon used the robot for the hardest screws and cut the time for that task from roughly 20 minutes by hand to about three. In Seoul, Hallym University Kangnam Sacred Heart Hospital opened a training centre on 18 September 2026 for robotic endoscopic spine surgery, which is done through two holes of about one centimetre each.

The same patient's spine after scoliosis surgery, straightened with a rod and screws, 2010. Photo: University of Utah Hospital Radiology via Wikimedia Commons (public domain)

ABB does not make surgical robots, and none of its arms is cleared for spine surgery. Where its robots could matter is the supply chain behind these operations. Screws, rods and cages have to be machined and finished in factories, which is routine work for industrial arms. A compact ABB IRB 1200 can tend a CNC machine cutting titanium screws, and a GoFa CRB 15000 cobot can load parts next to a technician. Bigger arms such as the IRB 6700 already grind metal parts in cells like Teqram's EasyGrinder. No source reports ABB doing this for spinal implants, so treat it as a possibility. Cells like these are usually laid out first in RobotStudio.

The new generation ABB IRB 1200, launched in 2025, a compact arm for machine tending and small parts. Photo: ABB

What comes next for robotic spine surgery

Spine platforms are moving from screw guidance toward planning the whole operation. Newer systems put the plan, the imaging and the robot in one package, and companies are adding software that tracks how the spine moves during surgery. Endoscopic keyhole approaches, like the one taught at Hallym, lean on this guidance because the surgeon sees less of the anatomy directly. Over the next few years more of this work may move to outpatient surgery centres, where a compact robot and a mobile CT scanner have to fit into a smaller room.

An ABB GoFa cobot loading parts into a machine at a small factory. Photo: ABB

Cost is still the biggest barrier. A spine robot with its imaging and disposables is a large purchase, and published comparisons have not found clear differences in pain or function scores at one or two years. Hospitals justify it with accuracy, fewer revisions, less radiation for staff and the ability to take on harder cases. The debate is the same one around knee replacement robots: the machine is more precise, and proving that patients do better takes large, long trials.

An ABB IRB 6700 grinding a steel part in Teqram's EasyGrinder cell at Ancofer. Photo: ABB

Skills are the other open problem. Surgeons still need to place a screw by hand when registration fails or the robot cannot reach, and training centres now teach both. At home, patients feel the change indirectly, through smaller incisions and, in some cases, shorter hospital stays. In factories, more demand for implants means more machining and finishing, where industrial arms from ABB and others could take on the repetitive, precise steps. For the longer story of how surgical robots began, see our post on surgical robot history.

An X-ray of a long spinal fusion with rods and pedicle screws from the upper back to the lower back. Photo: Nevit Dilmen / Wikimedia Commons (CC BY-SA 3.0)

Innovation

AI solutions for effortless ABB robot control.

Automation

Robotics

ceojohntran@chatwithrobot.net

+84905311611

© 2025. All rights reserved.

qtran1215@gmail.com