Since returning to Leicester, and getting a new phone, I’ve got back into the habit of tracking my movements and uploading gpx files to OpenStreetMap.

GPS traces give a rough idea of what the road layout looks like. For example, this is Artyom, a town about two hours north on the train from Vladivostok.


You can clearly make out the main thoroughfares, plus a few of the smaller capillary roads. The traces also give a good picture of the traffic flow.

This is what my commute looked like over summer last year, when I was regularly cycling out to Wigston.


And zooming in on the first part, here’s the route I used to take from student accommodation to the main campus.


Most of the main roads in Leicester have already got GPS traces. Despite this I continue to record my journeys, even on a route which I’ve already covered back and forth many times.

There are a couple of really interesting methodological issues which I’ve been thinking over in connection to this:

  • What’s the value in re-mapping ‘brownfield’ areas - going over existing data? How do contributors treat this differently to going forth into uncharted territory?
  • How do we measure map accuracy? At the moment the main tools for armchair surveying are satellite images + GPS traces. How should map contributions be balanced to prioritise accurate measurements?
  • How do we see roads on a map? What exactly is a road? Is it a line connecting a network, or is road surface a territorial feature in its own right? At the moment there’s a compromise, where you can tag road width and surface material, but the underlying data model continues to treat a road as a simple line.