IMPORTANT INFORMATION FOR TRACKERNET USERS - Change to data format

We want to let you know about a change to the Trackernet Detailed feed (AKA Feed 5) which may require updates to apps using this feed. The change goes live on 20th February 2019.

This change affects the feed which is available via both of the links below:
http://cloud.tfl.gov.uk/TrackerNet/PredictionDetailed// https://tfl.gov.uk/TrackerNet/PredictionDetailed//

The Trackernet internal system currently displays the actual train’s head code under the LCID field (Leading Car ID). However, in the Trackernet XML feed, the unique Train ID is incorrectly listed under the “LCID” field for the same train. After this change, the LCID field in the Trackernet Detailed XML will contain the actual train’s head code (which is unique within that data feed but there could be another train with the same LCID from another feed). There will be a new field for TrainID which will have the value that’s currently in the LCID field. You should amend your apps to ingest the new TrainID field rather than the LCID field. Please see attachment screenshot and sample new XML to enable you to test your apps in advance of the change.
TfL LCID Requirements Illustration.docx (507.5 KB)
ReturnForcastTimeXML.XML (8.5 KB)

We try to avoid make breaking changes to our open data but the operational and customer benefits to doing this meant we had to go ahead. The primary benefit of using the LCID field (as “Leading Car ID”) is to identify the head codes for Network Rail Trains. This information will enable us to identify the stopping patterns for Chiltern line trains which operate between Harrow and Amersham and allow us to provide more accurate predictions. This change also enhances the quality of our open data - you will now have access to the train numbers for all Network Rail and London Underground trains in the Trackernet feed.

Please post here if you need any further assistance with managing the change in your apps and we will do our best to help you.

4 Likes

I don’t think anyone uses that attribute anyway. The LCID was never the actual LCID of the train anyway so was just useless data. Still is. Though some developers may be parsing and displaying that data it’s purely as they don’t actually know its importance or lack of in this case, and are just display it expecting the end user to know its meaning. As both developer and tube driver I can say the attribute is pointless. Is it worth potentially breaking apps that do use the data for what essentially would be no significant benefit to anyone?? You could probably make the data a lot smaller by removing useless attributes like LCID and now the. New TrainID. I can’t see how it would have any benefit over the SetNo for developers or end users. If you are to change the feed actual LCID would help.

Thanks @ryan.forsyth for your feedback. I will pass this back to the Trackernet product manager to inform any future developments.

Hi. Today the Jubilee line trackernet went down. I know this cos I was driving one at the time. The data feed for the line though continued to show the snapshot of trackernet as if it was running fine. It wasn’t and all info was an hour behind reality. This really needs a way to be fed to the end user. It’s probably better for the feed to be down than showing incorrect information. This needs to be looked at please. It’s just not a good system where errors are not reported to the feeds end user. I have duplicated this post on yammer in the hope it’s picked up by your team.
Thank you