School of Nursing, Peking Union Medical College, Chinese Academy of Medical Sciences (CAMS-NIR Lab, Room 407)
 

The Nursing Physical AI Laboratory pursues a single guiding question: how can artificial intelligence move beyond understanding nursing to perceiving, reasoning, and acting within real-world care. Our work traces a complete pathway across the three layers of AI in nursing—generative, agentic, and physical—turning data into intelligent agents and, ultimately, into action in the physical world.
 

At the perception and documentation layer, the lab develops Voice ENR (智护声), an intelligent nursing documentation system that transforms nurses' language and clinical speech into standardized, structured data—forming the data foundation for everything that follows. At the cognition and decision layer, we focus on intelligent management of cardiovascular chronic disease: for the formation of exercise habits in populations with hypertension and high-normal blood pressure, we build a Health Agent integrating dual-process theory with reinforcement learning, while electroencephalography (EEG), functional near-infrared spectroscopy (fNIRS), and eye-tracking allow us to decode the neural mechanisms of cognitive–motor interaction, so that health interventions truly understand the person. At the physical AI layer, motion-capture systems record high-precision data from expert nursing maneuvers—patient turning, CPR, the Heimlich maneuver—to train a Nursing Large Action Model (Nurse-LAM), charting the path for nursing robots from understanding a task to performing the motion.

The lab is equipped with motion capture, EEG, fNIRS, eye-tracking, and 3D-printing platforms. Our achievements include granted U.S. patents, a Silver Medal at the Geneva International Exhibition of Inventions, and multiple national and institutional research projects, sustained through deep international collaboration with partners such as Seoul National University Hospital and Columbia University. By integrating nursing expertise, behavioral data, and Physical AI technology, we aim to build the essential training data and methodology for the coming era of nursing robotics—advancing the discipline of nursing toward new frontiers.