Data Verification Process

To be more precise, there is no vetting process, i.e. the data goes live as soon as the edit is saved. Of course anybody is free to create, for their own application, a snapshot of the data or a set of map tiles, and have them vetted by someone at their discretion.

There are some, optional, verification processes that kick in one an edit is saved and appears on the map. Most of these processes are manual and depend on monitoring tools; in some areas, local mappers actively monitor their "home turf" for changes and fix any newly introduced problems, but in other areas, bad data can exist for years without being detected. Some processes are automatic and will alert experienced mappers to large-scale modifications. All this is voluntary however; the quality control tools are largely maintained by members of the community.

The verification process is community based and starts when an edit to our data is prepared by the editing user. Edits are made to the live database without a verification (I guess there are some coarse automated checks for heavy vandalism in place on our servers), the users have to check their own edits before they make (save) the edit. Edits are usually prepared locally on a users computer/device - see our docu wiki page about "Editing". Also see my answer to a related question: how-soon-does-a-street-added-to-osm-appear-on-public-maps.

After an edit is made it can be viewed/reviewed by everyone in nearly real-time (depending on the latency of the tools used) and corrected (that includes "reverting") by every user if necessary (e.g. vandalism). That is a largely unplanned/-organised (in a classical way) community process. There are various tools available which can be used to analyse our data or to get notifications about specific edits/changes (see a blog article by me, wiki page "Quality assurance"). Most tools are run/queried on demand by a user. Other tools scan for possible problems and provide notification services to users who will then check the issue. Some few tools operate automatically (so-called bots) and fix minor and clear errors (e.g. Wall·E).

Please use the "add a comment" link below this answer to comment on this specific answer. If you want to add clarifying info to your question please use the "edit" link below it.

There are tools which help users to do some quality assurance*, but they are run manually by users on live data, so there is no formal, automatic error checking.

(*) for example OSMI