MetroMind is a transit intelligence project that learns how long NYC buses actually take to travel between specific stops.
Instead of trusting unreliable schedules, it observes real MTA bus data over time and builds a memory of how long each exact stop-to-stop segment usually takes at different days and times.
The goal is to use past behavior to make more accurate predictions in the future.
- Set up a script or service that connects to the MTA live bus data feed.
- Run the data collector continuously (24/7).
- For each bus, record:
- route ID
- previous stop
- next stop
- day of week
- time of day
- time taken to travel between the two stops
- Store all records in a database.
- Let the system collect data for an extended period (about 2 months minimum).
At this stage, the project runs quietly in the background and does not require a user interface.
The MVP is considered done when:
- The system successfully collects live bus data for the entire city
- Stop-to-stop travel times are reliably recorded for all routes
- At least ~2 months of historical data exists
- The system can output statements like:
“M5, Stop A → Stop B, Tuesday at 5 PM usually takes ~2.8 minutes”
- Data collection continues automatically after predictions begin
A mobile app, maps, or trip planning are not part of the MVP.