With so many contributors to OSM, it's possible to see inconsistencies in the way things are named and tagged. 

For example, there's at least one very prolific contributor on the Isle of Wight whose practices are a bit different to what we see elsewhere, and our journey planner just doesn't recognise how they edit things as being routable for a bus. 

These errors propose too many individual changes to be corrected and even if we did make the correction changes to suit a journey plan, the OSM contributor has the license to reverse the changes and "correct" them back again at some point. 

We have an automated pipeline for OSM data that takes the UK-wide OSM data, splits it up into smaller files, and then applies certain transformation steps to that data. 

We identify anything tagged in troublesome ways and rewrite the tags to what we would have expected such as the name of a specific, for example. That means we are able to produce the results we want and we don't have to battle against other people editing OSM.