Back to work Case study · UX / MedTech

ATS — Nightlife Triage

A medical UX concept designed for high-pressure nightlife emergencies in Dublin. ATS supports non-clinicians through the first critical minutes of care, when uncertainty and hesitation often delay life-saving action.

ATS Nightlife Triage cover image
Year 2025
Role UX / CC
Position Lead
Nightlife scene context

The Problem

In Dublin, nightlife is great… until it’s not.

  • Medical emergencies in social settings are common, but bystanders rarely feel prepared to intervene.
  • Hesitation during escalation can directly affect outcomes.
  • In Ireland, over 80% of drug-poisoning deaths involve mixing, while only 5% of people are first-aid trained.
User testing for ATS

The Moment

How can we help ourselves to help others?

  • Existing resources (Talk To Frank, HSE guidance) are largely scroll-and-search and require users to self-diagnose under stress.
  • Calling 999 is critical, but it does not provide stepwise, visual, in-the-moment guidance for immediate action.
  • In early testing, users showed no consistent understanding of recovery position before support was provided.

Solution Concept

Where does ATS step in?

User finds patient unconscious
User finds patient unconscious
User self-assesses
User self-assesses
User opens ATS to assess the situation
User opens ATS to assess the situation
App prompts and guides recovery position
App prompts & guides recovery position
User monitors situation in app
User monitors situation
If situation worsens, app escalates: ambulance, CPR guide, and emergency alert screens

If situation worsens, app escalates

Prototype

Demo video

System logic diagram

System Logic

Assess. Treat. Seek.

ATS mirrors real triage behaviour with a clinically ordered care flow.

  • Assess — prioritised questions identify risk quickly and reduce decision paralysis.
  • Treat — action-first, step-by-step guidance supports immediate care delivery.
  • Seek — the system escalates to emergency services when required and continues monitoring.

Design Principles

  • One question, one step at a time.
  • Decision authority for non-clinical users.
  • Action first, explanation second.
  • Designed for impaired and high-stress use.

Validated across 12 user tests (3 in person, 4 paper prototype, 5 digital) in simulated college, home, and nightlife contexts.

Outcomes

Why the project matters.

User testing context for ATS

Reduces bystander hesitation in medical emergencies

ATS interface supporting citizen-led response

Makes first aid response a more citizen-led effort

Nightlife safety context

Creates a safer response system in nighttime social settings

Implementation

Tech & tools.

ATS Chatbot: built with Cursor AI, ngrok prototype, backend with semantic search, HSE indexing, OpenCV, MediaPipe, ChromaDB, and Python dependencies

Research

What I found.

12 User Tests

3 in person · 4 paper prototypes · 5 digital prototyping

User tests conducted in simulated college, at home, and in a nightclub.

Interviews

“I found myself to feel a little bit, not only less safe, but less like people would intervene to help me if something were to happen.”

Existing Examples

  • AI triage agents dispatched in the UK to cut waiting times.
  • Red Cross First Aid only helps people through scroll guides and assumes they can diagnose the situation themselves.

Up next

Assemb3D — VR & web learning

View case