How it Works – Airport Delays

Airport delays API endpoints allow answering the questions, like How many flights are delayed or canceled right now or in past? or What is the delay statistics and delay index at an airport now or back then? These endpoints are designed to get a statistical insight into the overall “smoothness” of flight operations in this or that airport. They include:

Information returned from the endpoint routinely includes, per departures and arrivals separately, a median delay of recent flight movements, number of canceled flights, and flight delay index and validity periods of the batch.

Please review the links provided above for more information.

Considerations and Limitations

For each piece of delay information, we create a delay record. It may be transient or can be stored in a database depending on the nature of the request. In any event, each batch corresponds with a response you get for any of these endpoints.

  1. Delay records are based solely on the flight data we aggregate. Henceforth, these statistics inherit all the coverage and quality limitations applicable to the source flight data. This implies, among other things, that it is available only for those airports and those periods of time when delay information could have been reasonably derived. This concerns flights that:
    – qualify for this calculation (actual time and actual status of these flights must be available);
    – are arriving to or departing from the airports with Live or ADS-B + Scheduled quality.
  2. Delay records are calculated at 15 minutes intervals based on flight movements commenced in an airport within the 2 hours preceding to the beginning of the interval.
    E.g.: If you request delay information at London Heathrow for yesterday @ 09:23 AM, you will get the information calculated for flights commenced between 07:15 AM and 09:15 AM.
    Respectively, requesting the information for 09:30 AM, 09:31 AM, etc. (up until 09:44 AM) will give you the information calculated for flights commenced between 07:30 AM and 09:30 AM.


    Please note that historical data is calculated with varying latency. If you request delay info for the present or a very recent moment, you may get a transient (generated in real-time) delay record or you may as well get the latest available stored delay record. In any even, keep an eye on timestamps specified in FromUtc, ToUtc, FromLocal, ToLocal properties of the response payload.
  3. In order for certain delay information, like median delay or delay index, to be available, there need to be at least 5 qualifying flights in the batch. With lesser amount this information will be unreliable and won’t make much sence.
  4. Similar to many other solutions you may find out there, the delay index is valued from 0.0 to 5.0. The more flights in a batch delayed or canceled the worse this number will be. Calculation of the index is a bit convoluted, but it may essentially be perceived as a “magic” number telling you approximately how “bad” things are. The rule of thumb is that when 50% of flights in the batch will be delayed for 90+ minutes (or canceled), the delay index will be maximum.

Related posts