No time to recharge? Swap the battery. Worried about specimen tampering during transport? Put it in a secure storage chamber. Hon Hai’s Nurabot nursing robot overturns the traditional manufacturing mindset of the electronics industry by embedding user-experience and industrial-design teams directly into hospital nursing stations to observe and listen to nurses before designing. In Hon Hai’s smart healthcare strategy, what role does this robot play?

“Just like Takeshi Kaneshiro listening carefully to Donnie Yen’s breathing in the movie—we’re doing the same!” said Chiang, Chih-Hsiung, General Manager of Hon Hai’s Business Group B, gesturing toward the humanoid robot beside him. “Our team closely observed every movement nurses make. Only after spending a long time on-site did we come back and design this robot.”

Across Taiwan, nurses are the backbone of clinical operations—but the country now faces the most severe nursing shortage in decades. As hospitals struggle, Taiwan’s tech giants entering the medical sector are asking: in an era of low birth rates and rapid aging, how can technology reduce nurses’ workload?

Hon Hai’s answer is the Nurabot nursing robot. It not only marks Hon Hai’s formal entry into the healthcare services market but also reflects its deep user-experience research and strengths in product design.

Nurabot is the company’s first major attempt to “test the waters” in a new field as an EMS (electronics manufacturing services) leader. Developed in collaboration with Taichung Veterans General Hospital and Kawasaki Heavy Industries, the robot was redesigned entirely around the daily workflows of the hospital’s nursing station.

According to Chiang, since Chairman Yang-Wei Liu took office, Hon Hai has pushed to evolve from “Hon Hai the manufacturer” to “Hon Hai the technology company,” aiming to transform through AI and deliver comprehensive smart-living solutions. In Liu’s “3+3+3” strategy, the three selected industries are electric vehicles, digital health, and robotics—Nurabot being an embodiment of both digital health and robotics.

Once the decision was made, Chiang’s team faced the challenge: how to build a nursing robot that would truly work? Coming from the electronics sector, the team—though diverse in industrial design, algorithms, and medical management—knew little about hospital operations. They decided to abandon all assumptions and begin by observing nurses firsthand.

“I was a ‘little white rabbit entering the jungle’ too, so we had to go onsite. I didn’t want secondhand information—I wanted objective descriptions of what users really need,” Chiang said. “Normally, you send sales or engineers on first visits. This time we sent UX and industrial design experts to Taichung VGH to observe and experience.”

One month embedded in the nursing station, gathering insights from over a hundred nurses

Once stationed inside the hospital, the Business Group B team began an in-depth field study. “You have to deeply understand user behavior to identify bottlenecks and pain points. Observing for just one day is never enough. You need one to two weeks, sometimes a month,” Chiang said. This approach is rare in Taiwan’s electronics industry. “We were essentially behaving like a brand company. Most EMS companies never do this.” The experience shifted not only their methods but the entire design-thinking process of the product.

The team discovered that nurses spend a significant portion of their time on repetitive tasks. “Up to 30–40% of their work hours are spent walking back and forth between the nursing station and patient rooms,” Chiang explained, “because they must deliver specimens, medications, blood bags, and more.”

Beyond delivery tasks, nurses must repeatedly educate and brief patients. “Whenever a patient is admitted, the nurse explains the environment and health education—over and over again. Add ward rounds, noise monitoring, fall detection—it’s all highly repetitive work.”

After more than a month onsite and feedback from hundreds of nurses, the UX and design teams returned to Hon Hai to collaborate intensively with engineering and manufacturing teams on features and functions. The result: a robot very different from typical humanoid robots.

Real-world insight and negative feedback shaped the design

Nurabot’s appearance differs in key ways. First, the front and rear storage compartments: “In the original design, the robot carried specimens openly. Nurses immediately objected—they feared tampering,” Chiang said. “So they asked for an enclosed compartment.” This led to Nurabot’s sealed front and rear storage, with the rear serving as a refrigerated chamber for ice packs, blood bags, and temperature-controlled medical materials.

Biometric authentication is required to open the compartments. “If a specimen is assigned to Nurse Wang, the storage only opens with her biometric ID. If someone else tries, the robot verbally warns them not to interfere. These are all design-thinking results.”

Charging design was also reshaped by nurse feedback. “Nurses rejected a charging-based model,” Chiang said. “The robot must operate 24/7 and cannot pause to recharge. They need quick-swap batteries. That shows just how intense nursing workflow is.”

“Four Clouds” provide the strength behind the robot

Hon Hai’s advantages extend far beyond hardware. “Our strength is in the cloud,” Chiang emphasized.

“Nurabot is supported by four cloud systems—like a four-part harmony.” After design completion, the robot is first trained at Hon Hai’s internal AI training center, FoxBrain. Then it is simulated and tested in the digital-twin environment built with NVIDIA (Omniverse). On the robot itself, edge computing validation runs through NVIDIA’s Jetson AGX Orin. Finally, the robot integrates with the hospital’s private-cloud HIS system.

Behind Nurabot is an entire distributed AI infrastructure that must synchronize flawlessly. “At the FoxBrain center alone, we used 120 H100 GPUs—that’s NT$2–3 billion worth of computing power,” Chiang said.

Toward mass deployment and a centralized control platform

Nurabot is currently undergoing stress testing at Taichung VGH. After several months of real-world operation, the team has already identified hundreds of issues and is actively improving them. Through this testing, Hon Hai has gained valuable insights.

“We found that in any hospital, around 80% of nurse–nurse and nurse–patient interaction patterns are similar, especially at the nursing station,” Chiang said. “So if we create one successful product, we can replicate it across hospitals.”

More importantly, as more robots are deployed, Hon Hai will advance to the next phase. “When a large number of robots enter nursing workflows, they must communicate and coordinate through a central platform,” Chiang explained. “We call it the Central Flight Deck. Hon Hai is very good at managing platforms like this.”

Just like lanes in a city—bus lanes, sidewalks, fast and slow lanes—Hon Hai excels at orchestrating orderly movement among all robots. This is the foundation of the smart hospital, and beyond that, the smart city.

If you need, I can also prepare a concise media-edit version or a headline-focused investor summary.

Resource: 用30億元算力做一隻護理機器人！鴻海3+3+3策略，從蹲點護理站開始