As the world faces a looming shortage of 4.5 million nurses by 2030, according to the World Health Organization, Taiwan is offering a glimpse of how technology could help close the gap. In partnership with Foxconn and NVIDIA, Taiwanese hospitals are turning to AI-powered robotics to ease the burden on healthcare workers.

Figure 1. Nurabot

At the heart of this transformation is Nurabot, the world’s first collaborative AI nursing assistant. Designed to take on physically demanding and repetitive clinical tasks, Nurabot aims to reduce burnout and allow human nurses to focus on more meaningful aspects of patient care. Figure 1 shows Nurabot.

But Nurabot is just the beginning. Foxconn has developed a full ecosystem of smart hospital solutions built on NVIDIA’s AI platforms. These include models that monitor patient vitals, AI-powered clinical tools, and digital twin simulations that optimize hospital design and operations.

From Code to Care: A Digital Hospital Revolution

The process starts in powerful data centers, where large-scale AI models are trained using NVIDIA supercomputers. Hospitals then use digital twins — virtual replicas of wards and workflows — to simulate and refine their systems. Once tested, these AI capabilities are deployed to the real world via edge devices.

Top medical institutions across Taiwan, including Taichung Veterans General Hospital, Baishatun Tung Hospital, Mazu Hospital, and Cardinal Tien Hospital, are already integrating these technologies into their operations.

“Taiwan’s advanced healthcare infrastructure and commitment to digital transformation make it a prime environment for robotic integration,” said Shu-Fang Liu, deputy director of nursing at TCVGH, one of the hospitals trialing Nurabot. “Robots are helping us enhance care by allowing nurses to focus where it matters most.”

Powered by NVIDIA’s Jetson Orin, Holoscan, and the Omniverse platform, these AI systems are already transforming diagnostics, operations, and care delivery.

Among the standout innovations are:

• FoxBrain: A healthcare-focused large language model built with NVIDIA NeMo.
• CoDoctor AI: A clinical platform for real-time health monitoring, cancer screening, and more.
• CoroSegmentater: A coronary artery segmentation tool developed as part of the open-source MONAI project, improving diagnostic precision and surgical planning.

Digital Twins, Real-World Impact

Some hospitals are even creating digital twins of entire departments. TCVGH, for instance, built a virtual replica of one of its nursing stations, where Nurabot trained before entering service [1]. These digital twins allow hospitals to fine-tune layouts, simulate care scenarios, and train robotic systems in a risk-free environment.

Developed jointly by Foxconn and Kawasaki Heavy Industries, Nurabot is a blend of sophisticated engineering and advanced AI. It runs FoxBrain, trains in NVIDIA’s Isaac for Healthcare simulator, and uses Holoscan and Jetson Orin for real-time sensing and decision-making.

In practice, Nurabot handles tasks like delivering medicine, patrolling wards, and guiding visitors—potentially reducing nurse workloads by up to 30%, according to Foxconn.

“At one of our wards, Nurabot delivers wound care kits and patient education materials directly to the bedside,” said Liu. “It saves nurses countless trips to supply rooms, reduces physical strain, and frees up time for direct patient care.”

Nurabot is also proving useful during busy visiting hours and overnight shifts, when staff is limited. Looking ahead, nurses hope future versions will offer multilingual support, facial recognition, and even assist in lifting patients.

As Taiwan continues to lead the charge in AI-driven healthcare, its hospitals are offering a powerful model for how technology can help ease the global nursing crisis—one robot at a time.

Reference

1. https://interestingengineering.com/health/foxconn-nvidia-drive-ai-robotics-innovation

Cite this article:

Keerthana S (2025), Taiwan Deploys Nurabot, the World's First AI Nurse, to Address Healthcare Challenges, AnaTechMaz, pp.193