Evacuating health-vulnerable people
planned People who rely on powered equipment, dialysis, oxygen or intensive care are slower and riskier to move. This case study sets out how the Atlas could help, before any code exists.
Where it starts
Incident 5 of the Noto replay, "hospital power out, evacuate ICU patients", is the first test case. Today the Atlas can route that message; it knows nothing about who needs what kind of help.
What field input suggested
- Use the local government's electronic health records as the focal point, where they exist.
- Pilot in countries with strong patient-safety systems first, then extend.
- Consolidated digital health records are the precondition. Today the data sits in silos.
Proposed design planned
- A health-vulnerability layer beside buildings, roads and lifelines.
- Consent-based and aggregated: counts or flags per area from the record owner, never raw records.
- Behind the same zero-trust gate, plus privacy law (Japan APPI, Taiwan PDPA and similar rules).
- Near-zero extra work for the people who maintain health records.
- Pilot in Japan using existing municipal evacuation-support registries, then a second developed system.
Open questions
- Who owns the data feed in each pilot country, and what is the legal basis for sharing it in an emergency?
- Which conditions matter most for evacuation timing? Clinicians, not engineers, should decide.
- Can the layer work offline on the prefecture nodes when the central link is down?
If you work in clinical care, municipal disaster planning or health privacy law, this is where help is most needed.