We extract all the OpenStreetMap objects touched by our team, for the previous week.
Each team member reviews the changes for accuracy, use of appropriate tags, and relevance to the current mapping project.
If a reviewer sees any errors, they add a comment in the respective changeset discussion, for fixing immediately.
Automated error detection
Going forward, we are strengthening the peer review process using automated error detection, to instantly flag mapping errors to the team and post them to our Slack channel. Leveraging from the scalability of TileReduce with OSM-QA-tiles, we are openly helping to develop OSM Lint to process common data errors at a global scale.
It would be great to hear your feedback on our process and how we can make it better. Feel free to drop by our mapping repository and post your suggestions or get in touch with me on twitter.