WayCare

Login for the website
Hospital layout
Route pathing example

Inspiration

Hospitals move patients constantly, but routing still happens by phone calls and guesswork. We’ve all watched rooms sit idle while everyone waits. We wanted a simple, deploy-today tool that makes patient moves predictable, visible, and fast.

What it does

WayCare locates any patient, lets staff pick a destination, and returns the fastest safe route with a live ETA-accounting for elevator waits, crowded hallways, blockages, and room readiness. It auto-re-routes when conditions change and surfaces shared ETAs so receiving teams can prep on time.

How we built it

Three.js visualizer for real-time rendering of patient location and routes on individual hospital models (scaffolded to scale to a full facility map and system-wide view). React frontend (fast iteration, component reuse) with status chips, ETAs, and live re-route to keep staff informed. Node.js backend + PostgreSQL for a clean API layer and persistent hospital + patient records (units, rooms, elevators, constraints, transports, audit). Routing engine in the backend: graph traversal on a hospital graph where points (junctions/doors/elevators/rooms) are nodes; edges carry time-dependent weights (base travel time + elevator wait + crowding + closures). Computes shortest safe path with dynamic re-routing for obstacles. Designed for scale: roadmap to multi-transfer optimization (capacity & priority aware), batching windows, and slow-down handling across multiple simultaneous moves. Planned integrations: WebSockets for live ETAs, optional RTLS/QR inputs for location, and later HL7/FHIR hooks for order-driven transports.

Challenges we ran into

Turning messy buildings into a maintainable graph (units/elevators/constraints). Stable ETAs with noisy, bursty elevator waits. Keeping privacy minimal while still being useful.

Accomplishments that we're proud of

The technical demonstration Being able to make the 3d hospital in blender

What we learned

In transport, visibility beats perfection-shared ETAs reduce chaos immediately. Small, reliable signals (elevator latency, historical peaks) drive big wins.