NVIDIA's Open Simulator Set to Transform Surgical Robotics

Before a healthcare robot can operate safely in a lab environment, it must first learn to interpret physical resistance. 

Given the sheer volume and variety of data developers need to train, test and improve robot behaviour, ingraining this knowledge into prospective robots creates one of the biggest bottlenecks in healthcare robotics.

Open source simulators like NVIDIA’s Medical Physics Simulation framework, which enable healthcare teams to access and build upon shared simulation tools and AI models, are crucial to unlocking these bottlenecks and accelerating industry-wide surgical robotics progress.

"The framework brings together anatomy and medical device behaviour with sensor simulation and robot learning so teams can create reusable simulation environments instead of rebuilding custom scenes for every workflow, saving developers time and bringing innovations to market faster," writes David Niewolny, Director of Business Development for Healthcare and Medical at NVIDIA, in a product announcement.

Enabling a “virtual training ground” for medical robots, development teams can run thousands of simulated procedures at once, helping them identify problems earlier and focus costly lab time on the most important real-world tests.

Medical robotics firms advance surgical simulation

NVIDIA's healthcare simulation technologies are already being adopted by leading medical technology companies, providing an early indication of how simulation could reshape surgical robotics development. 

CMR Surgical and Cambridge Consultants, part of Capgemini, are using Cosmos-H-Dreams, a real-time generative AI physics simulation capability in the Medical Physics Simulation framework, to create patient-specific surgical simulations by modelling the complex interactions between surgical instruments and soft tissue.

The collaboration has contributed almost 500 hours of anonymised clinical data, supporting research across procedures including gallbladder removal, prostate surgery, hernia repair and hysterectomy.

"Open source models allow us to build on shared knowledge, accelerating responsible innovation and, ultimately, give us the potential to deliver more consistent care and better outcomes for patients worldwide," says Chris Fryer, Chief Technology Officer at CMR Surgical.

Elsewhere, Johnson & Johnson MedTech is using NVIDIA's Medical Physics Simulation and Cosmos foundation models to develop digital twins of its MONARCH robotic platform, enabling the simulation of complex urology procedures and kidney stone treatments.

Other medical technology companies are also adopting the platform, with XCath training autonomous endovascular robots, Inner Logic generating synthetic data for device validation and Medtronic Structural Heart exploring simulated X-ray data for catheter navigation research. 

NVIDIA’s growing medical footprint

While the new simulation framework targets surgical robotics, it also reflects NVIDIA's broader healthcare strategy. 

Across life sciences, the company is increasingly positioning its AI technology as the foundation for drug discovery, medical imaging and clinical research. 

In July 2026, Bristol Myers Squibb expanded its partnership with NVIDIA to build what it describes as the "most powerful AI factory in life sciences" to accelerate drug discovery and foundation model development.

At NVIDIA GTC 2026 in March 2026, Novo Nordisk was announced as an early pharmaceutical partner for NVIDIA's Proteina generative AI model, with Novo using NVIDIA BioNeMo foundation models to accelerate structure-based drug discovery.

Partnering with innovative AI startups

NVIDIA's healthcare ecosystem is also expanding into medical imaging through strategic partnerships. 

Following the integration of NVIDIA's NV-Reason and NV-Generate foundation models into the HOPPR AI Foundry earlier this year, HOPPR has now introduced its own Chest CT Narrative Model, extending the platform's growing portfolio of imaging-specific foundation models. 

The announcements demonstrate how NVIDIA's accelerated computing and open AI models are enabling partners to build specialised clinical applications across healthcare. 

“Chest CT is one of the most information-dense studies in radiology. Getting AI to work well across everything it captures – the lungs, the heart, the aorta – is a genuinely hard problem, and we are pleased with what this model can do,” says Khan Siddiqui, Co-Founder and CEO of HOPPR.

“But the model is only part of what we are building. Our AI Foundry and Forward Deployed Services include secure infrastructure, curated data, and the clinical and technical expertise to help any organisation move from a foundation model to a working application, regardless of where they are starting from. That is what makes this more than a model release.”