The Island of Healing Machines: Inside Taiwan’s Bold AI Healthcare Experiment

Foxconn, Nvidia, and Kawasaki are unleashing AI-powered nursebots to ease Taiwan’s nursing shortage — and redefine care for an aging world.

Jun 13, 2025

Image credit: NVIDIA blog. The chest of the robot reads: “Jensen Huang loves me.”

It’s 2 a.m. at Taichung Veterans General Hospital.

While human nurses move between patient rooms, one of their newest helpers quietly navigates the brightly lit hallways: a mobile service robot designed to carry out repetitive but essential tasks.

The robot — called Nurabot — moves autonomously, delivering medications, transporting lab samples, and guiding visitors through the hospital’s complex layout. With its sensors and AI-powered systems, it avoids obstacles and ensures supplies reach the correct locations.

Nurabot is part of Taiwan’s bold effort to relieve overwhelmed healthcare staff by automating routine hospital work.

The Problem: A Growing Nurse Shortage

Across the world, healthcare systems are straining under staffing shortages. The World Health Organization predicts a shortage of 10 million healthcare workers by 2030. Taiwan, with one of the fastest-aging populations globally, faces the challenge sooner than most.

Night shifts, heavy patient loads, and repetitive tasks contribute to nurse burnout. By taking over tasks that don’t require human judgment or emotional care, robots like Nurabot give nurses more time to focus on direct patient care.

“Robots like Nurabot are augmenting the capabilities of healthcare staff, enabling them to deliver more focused and meaningful care to patients.” — Shu-Fang Liu, deputy director of the nursing department at TCVGH

The Technical Lowdown

For readers curious about how Nurabot operates, here’s a peek under the hood:

Nvidia Jetson Orin: The onboard AI processor that allows Nurabot to make quick decisions and navigate in real time.
Nvidia Holoscan: Processes sensor data to interpret surroundings and patient information.
Isaac Sim (Nvidia Omniverse): A simulation platform where developers train and test Nurabot’s navigation and task-handling skills.
FoxBrain (Foxconn’s language model): Supports Nurabot’s ability to interact with staff and patients through speech recognition and natural language processing.

This technology allows Nurabot to perform tasks like delivering medications, transporting medical specimens, monitoring vital signs, and guiding visitors through hospital facilities.

Building the Smart Hospital Ecosystem

Nurabot is only one part of a much larger healthcare transformation underway in Taiwan.

Foxconn, Nvidia, and the Taiwan government are building an AI Factory — a massive supercomputing center equipped with 10,000 Nvidia Blackwell GPUs. This facility will support:

Digital twins for hospital logistics and facility planning.
AI-powered patient monitoring to predict health risks earlier.
Training simulations for medical staff and robots.
Ongoing development of advanced hospital automation systems.

The aim: to create highly efficient, AI-augmented hospitals that can better serve growing patient populations with fewer available staff.

You can see the bot in NVIDIA’s video on AI healthcare.

Why Taiwan’s Approach Matters

Taiwan’s experiment could serve as a model for other countries facing similar healthcare challenges. Success could show how AI and robotics can safely support human caregivers. Failure could expose the limits of automation in fields that depend on personal interaction and empathy.

Keeping the Human Touch

While Nurabot handles physical tasks, hospital administrators emphasize that it’s not designed to replace nurses. Medical ethicists caution against losing sight of the importance of human contact in caregiving. For now, hospitals are using Nurabot to handle routine duties so that nurses have more time for meaningful patient care.

Vocabulary Key

Edge Computing — Processing data directly on the device rather than relying on external servers.
Digital Twin — A virtual replica of a real hospital used for planning and training.
Holoscan — Nvidia’s platform for processing real-time sensor data in healthcare settings.
Isaac Sim — Nvidia’s simulation platform where robots learn to navigate and complete tasks virtually.
Large Language Model (LLM) — AI trained on massive text datasets that can understand and generate human language, like FoxBrain.

5 Quick FAQs

Is Nurabot a humanoid robot? No. Nurabot is a mobile service robot designed to perform hospital tasks autonomously.

2. Where is Nurabot currently being used?

It’s operating in several hospitals in Taiwan, including Taichung Veterans General, Baishatun Tung, Mazu Hospital, and Cardinal Tien Hospital.

3. What kinds of tasks can Nurabot perform?

Medication delivery, specimen transport, vital sign monitoring, visitor guidance, and simple mobility assistance.

4. How much does Nurabot reduce nurse workloads?

Early trials report up to a 30% reduction in routine tasks during busy shifts.

5. What features may be added in the future?

Plans include multilingual support, facial recognition for patient interaction, and enhanced mobility assistance for patient lifting.