How do predictions work?

High-level, Swiftly Transitime generates an arrival estimate for passengers by doing the following: 

1. We use your historical vehicle position data to understand how fast vehicles move at certain times of day, on certain road segments, etc. We often analyze tens or hundreds of millions of historical data points to really understand how your transit network performs under different circumstances.
2. We then combine this historical data with real-time information we observe on the road.
3. Finally, with that combination of real-time and historical data, we generate an arrival prediction based on all the data points we observe. We then automatically regenerate the prediction every time we receive a vehicle position update, ensuring that passengers always have the most up to date predictions.

Once a prediction is made, it can be sent anywhere through data feeds and APIs. Swiftly Transitime can send data to mobile apps, web pages, automated SMS systems, automated voice response systems, electronic stop displays, and more.