[OSM-talk] Adding Missing Roads in the Caribbean

[Jeroen Hoek]

On 01-02-2022 09:58, Alan Mackie wrote:
> How have these missing roads been validated? In some low lying areas 
> scars that look like roads can persist for decades in imagery despite 
> having become entirely unsuitable for travel.

Good question, but TomTom is loath to share any details of their work. 
They do announce their intentions (they have been posting updates in the 
Dutch community subforum for the Dutch roads), but Marjan does not 
appear to have clearance to go into details about their source and 
methodology.

This being TomTom, a seller of navigation aids, our best guess is that 
they use the collected GPS-traces of their users (which they consent to 
in the privacy policy) to create an amalgamated heat-map, and match that 
with OSM ways given a width based on hierarchy and lanes. Any place the 
has traces but no OSM way get marked as needing attention.

We don't know for sure, because TomTom avoids answering any questions 
about this. They did write a blog post that shows their 'mapmetrics' tool:

https://www.tomtom.com/blog/maps/the-battle-for-quality-maps/

But how the sausage is made TomTom won't tell, which makes helping them 
fix false positives (and preventing well-meaning mappers from 'fixing' 
these using their MapRoulette challenges and a complete lack of local 
knowledge) quite hard.

For the cooperation with TomTom to be more beneficial to OpenStreetMap 
as well, it would really help to have access to their (anonymized and 
flattened) heat-maps, but it is a shame that this seems out of the question.
-------------- next part --------------
A non-text attachment was scrubbed...
Name: OpenPGP_signature
Type: application/pgp-signature
Size: 495 bytes
Desc: OpenPGP digital signature
URL: <http://lists.openstreetmap.org/pipermail/talk/attachments/20220201/78aac504/attachment.sig>