A mobility map does not come from a single database. It combines timetables, street geometry, live positions, fares and layers such as parking, fountains and charging points. Turning those files into a useful answer takes more than drawing them on a map.
Schedules and live data are different
GTFS Schedule describes networks, stops, routes, trips and service calendars. GTFS Realtime can add vehicle positions, arrival predictions and alerts. The first dataset is relatively stable; the second changes every few seconds and can go offline.
An honest product preserves that distinction. If a live feed fails, it should not turn a scheduled time into a false real-time arrival.
The street is a dataset too
OpenStreetMap provides streets, directions, access information and many services. Public portals publish traffic, air quality, beach status or fuel prices. Every source has its own coverage, frequency and licence. It needs an origin, validation and attribution.
In emap, a source enters a catalogue, is validated against schemas and becomes domain data. OSRM calculates road routes and OTP combines public transport. The API sends the same contract to web, mobile and the MCP server for agents.
emap/labs adds reproducible evaluation for geographic search. It includes Spanish and Basque cases and measures abstention too: acknowledging that evidence is missing is part of quality.