GTR vs GEMINI seating capacity discrepancies

[ciaran haines]

I'm looking at the capacities of trains for GTR by consuming the GTR passenger loading data from RDM and the GEMINI feed. I'm finding a lot of very large discrepancies - more than half of trains are more than 20% of seating capacity out, with the GTR seating capacity data usually being higher, and the number of carriages often being consistent.

GTR report seating and standing capacity separately so I'm fairly confident that the problem isn't merely chucking these figures together. These discrepancies are often very large, and it appears to hold across all of the GTR subsidiaries, including the Gatwick Express which I think only has one train class.

Does anyone have any insight into what's happening? I'm open to this being a failure of my understanding, but I'm at a bit of a loss. 

[Christoper Stafford]

Iiirc seating capacity figures in GEMINI are often wrong. In particular trains which have mixed first and standard class seem to only report the seat count from one of those classes. If the mixed first/standard carriages are only reporting the first class seats then that probably accounts for a 20% discrepancy. 

Gatwick Express trains are advertised as carrying first and standard class, and the GEMINI feed is reporting one of the carriages from each unit as only having 22 seats, which seems on the low side.

[Andrew Bolton]

You're correct that composite coaches don't have the correct seating in the feed - it only gives the number of first class seats and ignores the standard ones. For those that have access to the internal system Web Gemini, you'll see the same issue there. I did raise a ticket with Worldline for it a few months ago but I'm not sure it went anywhere!

You're also right that the underlying data isn't always great however I can assume you that GTR's data is correct in R2 which is the ultimate source for the data. The same may not be true for all TOCs though!

Andrew

[ciaran haines]

Ah, that will explain a chunk and gives me hope that I'll be able to shrink the number of unknown errors. Thank you both!!!