[OSM-talk] The completeness of OpenStreetMap

[Andy Robinson (blackadder-lists)]

Edward Johnson
>Sent: 25 June 2008 4:04 PM
>To: talk at openstreetmap.org
>Subject: [OSM-talk] The completeness of OpenStreetMap
>
>I'm compiling some research into ways in which we can think of a map as
>complete and how complete OSM is. I was just wondering if anyone had any
>thoughts on the topic. My initial ideas have been posted on my blog
>http://edwardmjohnson.wordpress.com/ and I will continue to update this
>until I'm satisfied I know everything there is to know about the
>completeness of the map!
>
>One of the most interesting parts is how can we quote completeness of the
>map? My first thought was a stage system as quoted below
>
>* Preliminary stage - GPS tracks gathered or area has yahoo imagery.
>
>* Stage 1 - Nodes and ways mapped onto OSM using GPS track or aerial
>imagery.
>
>* Stage 2 - All roads named and roughly categorised
>
>* Stage 3 - Map good enough for satellite navigation. All one-way streets,
>and restrictions tagged along with accurate street categorisation.
>
>* Stage 4 - All POIs (i.e. post boxes, bus stops, pubs, restaurants,
>supermarkets etc.) tagged.
>
>These stages may or may not occur in sequential order and each stage can be
>quoted complete in terms of percentages. For example we might say that
>London is 100% complete for Preliminary and stages 1 and 2, but only 60%
>stage 3 complete and 20% stage 4 complete. The hard question is how do we
>accurately measure these percentages.
>
>What are other people's ideas?


Unfortunately it doesn't really work like that. We might have a very small
part of a city that is stage 4 and then next door in the same city it might
not even have reached stage 1. That's why I had a stab at looking at it a
little differently and came up with
http://wiki.openstreetmap.org/index.php/Completeness_Metrics  but its not
that scientific and is only intended as a guide to help focus mapping
effort.

Cheers

Andy