A Safe Shift - VR x AI Patient De-Escalation

An immersive multi-user simulation for emergency protocol training. This VR solution enables students to navigate complex in-ward crises, fostering teamwork and psychomotor skills without patient risk. There were 4 scenarios based on Real Medical Incidents: ACLS and Defibrillation Safety, acute myocardial infarction, IV dislodgement and medication overdose and workplace violence.



Client

Faculty of Medicine, The Chinese University of Hong Kong

Year

2025-present

Link

https://nurseverse.nur.cuhk.edu.hk/

End User

Nursing Students

Technology

AI reasoning, VR, Machine Learning

The Challenge:
Clinical workplace violence is a global crisis, yet traditional training relies on scripted actors or static videos. These methods fail to replicate the volatile, high-stakes nature of a patient in crisis, where a single wrong word can shift a verbal dialogue into a physical confrontation.

The Solution:
Angry Shiba Studio engineeredA Safe Shift—a pioneering XR ecosystem that replaces scripts withDynamic Behavioral AI.

The Engineering of Interaction:

  • Context-Aware AI: We fed complex patient histories and behavioral triggers into the AI Case Nurse, allowing for a realistic clinical briefing before the interaction begins.
  • Real-time Sentiment Analysis: The AI Patient doesn't just "play a clip"; it listens. Using real-time voice-to-voice processing, the patient's aggression level fluctuates based on the user's tone, choice of words, and clinical empathy.
  • The Contingency Loop: Failure to de-escalate triggers a seamless transition into Workplace Violence (WPV) physical protocols. This ensures a holistic training cycle: from verbal intelligence to physical safety.

The Human Side of the Tech:"A Safe Shift" is an exercise in Empathy Engineering.By providing a safe, virtual space to fail, we allow nursing staff to build the emotional muscle memory required to stay calm in the face of aggression. At the end of every session, the system generates an AI-driven De-brief, providing precise data on where the LEAP model succeeded or failed.


The Impact:
We are vitalizing the safety of frontline healthcare workers. By combining neural intelligence with spatial immersion, we ensure that nurses are equipped with more than just protocols—they are equipped with experience.