As with any automated service, we have some limitations that are placed upon us by our data sources. Most of these do not cause an issue during the normal operation of a railway service, but we list the limits that we are constrained by below for your information.
Services report using a wide network of rail-borne position monitoring systems or using manual input by signallers, or other staff, into a system operated by Network Rail. We use these reports, and other information we collect, to generate predictions on the trains likely future progress.
The prediction engine we use is built to predict on the basis of the services's schedule and, additionally, the passage of previous services in an area. We use the reports generated by the monitoring points in order to improve the accuracy of our predictions.
Some locations do not report into our system. Once a train has been known to have passed through these locations, we replace the predicted times with 'No report'.
We show planned platforming information as is detailed in the timetable distributed from Network Rail on an advisory basis only. At some stations, they operate on a daily platforming plan which differs from that detailed.
We can 'confirm' platforms (when they go bold) based on the passage of presence of a train service. Therefore, it is possible that platform changes can be advertised on a station before we are aware of them.
In many areas, the confirmation on platforms based on the passage of a service works reliably. In some areas, it is possible that the passage of the train may create inaccuracies within the platforming data as this works on a probabilistic basis. Where this happens, we are alerted and will investigate - but, when it does occur, it is normally only a very short walk to the other platform.
We use systems which are intended as a representation of the train service on an 'after the fact' basis. We make great efforts to transform the data that we receive in order to be able to use these on a 'before the fact' basis.
We are wholly reliant on manual input into internal railway systems of cancellation information and, therefore, note the limitations of these systems below.
If a train is partially cancelled from an origin, we are normally able to find this out in advance. If a train is running and is cancelled en-route, we will not know until the train has reached this en-route station.
Some train operators update the timetable system with new schedules on a short term basis - this can only happen over an hour before a train is planned to depart, or if the previous schedule for a service is cancelled. In these instances, we will generally be able to provide information if a service is curtailed.