A student project
for a public problem.
GeoHealth Compass is a capstone-scale MVP that takes two questions seriously: "which hospital should I go to?" and "does my region have fair access to care?" — and answers them on the same map.
Why healthcare accessibility matters.
Finding the right hospital in Indonesia is harder than it should be. A person in urgent need might consult three different apps, a search engine, and a family group chat before making a decision — under time pressure and often under stress.
At the same time, the distribution of healthcare across the archipelago is uneven in ways that are rarely visible to the public. Provincial-level differences in hospital density, specialty breadth, and service capability are real, but they're buried inside reports that most people will never read.
GeoHealth Compass tries to do something modest but useful for both problems: put them in one interface, built with the same geospatial backbone, and explain the reasoning clearly.
Three pillars, one platform.
Geospatial discovery
Your browser shares your location only locally. We compute great-circle distance with the haversine formula and sort candidates by proximity, travel-time estimate, and service fit.
Transparent scoring
Every recommendation is a weighted composite of five factors — specialty, distance, rating, service completeness, and access. Emergency mode reweights toward urgency. You can see each factor's contribution.
Regional intelligence
Provincial data is aggregated into a Healthcare Access Index. Underserved areas are called out explicitly, with honesty about dataset limits.
How the recommendation works.
All factor scores are normalized to [0, 1] before weighting. Ratings from the 1.0–5.0 scale are linearly mapped to [0, 1]. Distance is capped at 50 km. Service completeness counts six standard indicators (24h ER, ICU, ambulance, BPJS, trauma, NICU). The final score is rescaled to 0–100 for readability.
Speech is the fastest accessibility affordance we have.
The voice feature uses your browser's built-in speech recognition to transcribe a short command, then maps keywords to the same filter and recommendation system available through the UI. It recognizes common phrasings in both English and Bahasa Indonesia.
Important: voice search is not a chatbot, triage assistant, or symptom checker. It is an accessibility-first input method for the same search you could perform by tapping filters — designed for elderly users, users with motor impairments, and users in urgent physical conditions.
What this project is — and isn't.
- A decision-support tool for hospital discovery
- A visualization of regional access patterns
- An accessibility-oriented voice search interface
- A transparent, explainable recommendation engine
- A student capstone MVP built for public value
- A medical diagnosis or triage system
- A hospital booking or appointment platform
- A source of real-time bed availability
- An authoritative clinical directory
- A substitute for calling emergency services (118/119)
Try the compass.
Start on the explore page, or read the numbers behind the map.