Improving OSMCHA

Posted by wille on 20 February 2017 in English.

This month I came back to work on development of OSMCHA, a project that I started in the middle of 2015, with the aim of helping to detect potentially harmful OpenStreetMap edits.

My motivation to start OSMCHA was that I felt that we needed a better tool to monitor changes made to OSM data. Looking at statistics of how many nodes were created, modified and deleted in a day in Brazil, it was common to see some peaks in the number of created or deleted nodes. This made me curious to know - where are those changes… Has an import been made? What is the source of the data? Has someone deleted an entire town?

Furthermore, sometimes we discovered changesets that explicitly said that the data came from Google Maps or some other inappropriate source. Despite that information being expressed in the changeset metadata, it was common for those edits to remain in the OSM database for months or years without anybody seeing them.

As it is impossible for the community to review all edits, my idea was to create a tool to point the changesets that potentially can damage the map and to analyse the metadata searching for suspect words.

I showed OSMCHA for the first time in the SotM Latam 2015, in Santiago, Chile. The Mapbox data team became interested and started using it. Moreover, they developed some new features and now I’m very excited to announce that Mapbox is sponsoring my work in the next few months to improve OSMCHA.

In the last two weeks we enhanced documentation, lifted test coverage, cleaned up the code and prepared the code-base to receive the updates we are planning. The main move we are planning to OSMCHA is to rebuild the interface and make it map-based (instead of showing the changesets on a list, we want to exhibit it over a map). We have some ideas to make it easier for users to review the edits in their area of interest, have some notification features and improve the analysis of changesets, among others…

Nevertheless, we want to listen the suggestions of the OSM community to build something better. If you have never used OSMCHA, try it on You can post your feedback and ideas here or open an issue on our github repository.

Finally, I would like to thank Mapbox for the support, specially Sanjay Bhangar, who since the beginning was excited about OSMCHA and believed in the potential of the project!

Read other posts about OSMCHA:


Comment from PlaneMad on 20 February 2017 at 14:26

Wille, your work has been phenomenal in laying the groundwork for a more powerful community validation tool on OSM. For instance: list of all changesets with no additions and atleast 5 deletions in London.

Really excited with all the developments so far and what the future of OSMCHA looks like with a map interface!

Comment from andygol on 20 February 2017 at 14:31

Thank you for the timely review, I’m just going to tell about OSMCHA this week at the GIS Forum in Kharkiv, Ukraine.

Comment from jengelh on 4 March 2017 at 11:29

OSMCHA output: “Source: not reported”

The source tag is often placed on the feature (node, way) rather than changeset, since a new building may be visible on one but not on another other source (time difference of imagery).

Login to leave a comment