Mapping railway worker accidents

No doubt you’ll all have read – avidly – every line of the 50,000 or so cases currently in our database. Well – perhaps not then. And no wonder. The density of information is an amazing thing, one of the aspects that makes what we’re doing so valuable. Whilst it’s of course possible to search the data, that’s most useful if you know what you’re looking for. Casual browsing is less easy. So, we’ve been thinking a bit about more accessible ways into the project data – including mapping. We’ve put together a couple of trials, including this one, for the Portsmouth area:

Why might this be useful? It’s visual, and easier to use and see. In the map, we’ve imported details from the database and then pinned each case to an approximate location for the accident. Clicking on the icon brings up the details from within the database, so we have all of the information – but with the spatial element clearly visible.

This shows us the railway geography – not just where lines are (or were) but how cases were distributed. We can start to see the railway’s relationship to its wider environment and local history, and start to explore those links in greater detail. And it might show us concentrations of cases, which might indicate particular problems or risks, for example.

We’ve always got to be careful to understand the underlying data too. Sometimes because of the information in the original records it’s not possible to pinpoint exactly where an accident occurred. In our example, the Portsmouth map draws upon the cases from the Railway Inspectorate data. The Railway Inspectors only investigated about 3% of all worker accidents – so what’s in map is absolutely not all of the accidents that would have occurred in the area. So far we’ve not included cases from the trade union dataset – that’s much less geographically fixed (as it was grouped by Branch, rather than accident location).

So, mapping like this isn’t a simple thing – but it’s still valuable, and something we’re keen to develop. We’d love to use old Ordnance Survey maps as a base layer, for example, to really show the landscape on the ground.

We’d welcome your thoughts on how you might want to use a tool like this. What would you want to see in it? What should we be trying? What questions should we ask? Do contact us and tell us!