An innovative system has been touted the first ‘surgical humanoid’ engineered for the operating room.
It is already a commercially deployed surgical humanoid and, with FDA clearance and clinical use underway, LEM Surgical plans to evolve Dynamis using NVIDIA Physical AI to reshape hard-tissue robotic surgery.
The Swiss-based company showcased the Dynamis surgical system at CES 2026.
Designed for spinal and orthopaedic procedures, it features a multi-arm architecture that mimics human bimanual workflows, allowing surgeons to position instruments with sub-millimetre accuracy.
The system leverages NVIDIA Isaac for Healthcare and onboard AI processing to compensate for micro-movements and maintain stability during procedures.
The Dynamis is an integrated, navigation-based robotic platform designed to support a wide range of spine surgical procedures, including thoracic, lumbar, and sacral approaches.
Built for use in real operating rooms, the system combines real-time tracking with advanced robotic guidance to enhance surgical accuracy and workflow efficiency.
In November, LEM announced the successful completion of the first clinical procedures in the US utilising its Dynamis system.
At the core of the system is a three-arm robotic architecture.
Two computer-controlled robotic arms assist with surgical guidance and precise instrument positioning, while a third robotic arm is dedicated to holding and controlling the navigation camera.
All three arms are integrated into a single mobile cart, partially positioned beneath the surgical table, minimising the operating room footprint while maintaining stability.
The cart incorporates bilateral elevator mechanisms that enable the robotic arms to be securely deployed, folded, and locked in place as needed during procedures.
According to LEM, the system is designed to seamlessly integrate with multiple intraoperative imaging platforms, using DICOM image data to support navigation and planning.
Its software suite manages motion control, navigation, user and case management, data storage, and safety functions, providing a unified control environment for surgical teams.
Dynamis has received FDA 510(k) clearance and is being routinely used for spinal procedures at Southern Hills Hospital in Las Vegas.
The company is now outlining a forward-looking strategy that positions NVIDIA’s Physical AI ecosystem at the centre of future development, aiming to redefine how hard-tissue robotic surgery is performed.
The firm now intends to integrate NVIDIA Isaac for Healthcare, an open-source framework tailored for medical robotics, along with NVIDIA Cosmos, an open-world foundation model.
Together, these platforms will enable continuous training and refinement of robotic capabilities, allowing the system to learn from data, adapt to varied clinical environments, and provide surgeons with augmented, high-precision assistance.
Yossi Bar, founder and CEO of LEM, said: ‘By advancing physical AI, we aren’t just building machines; we are architecting the future of human-robot collaboration to enhance clinical precision and expand access to life-saving care.’
