Microgrants/Microgrants 2020/Proposal/Address and Street Completion in Luxembourg

From OpenStreetMap Wiki
Jump to navigation Jump to search
statusproposed
Address and Street Completion in Luxembourg
Build tooling to enable the import and management of road and address data in Luxembourg
start-date2020-09-01
end-date2020-09-31
budget (Euro)1800
grant_typeINDIVIDUAL
location(s)Luxembourg
contact(s)• david.moraisferreira@gmail.com


About your Mapping Community

Tell us a little bit about your mapping community at the moment, some examples of details to include are below, but please provide us with any information you think is relevant. Please write 100-200 words

I have been an active contributor and an OSMF member since 2018. The community in Luxembourg is currently not centrally organized, although most recurring mappers know each other. I have created a Matrix chat room in the hopes of regrouping both new and experienced mappers.

What do you need the grant money for?

We want to understand what you plan to buy, how much you plan to buy, and why this will help you to scale your community mapping efforts. There are two tables for you to fill in your response, and you may add more rows as needed. Some examples are in the table for your reference. You are welcome to use a different format in your response, but please make sure to include these details.

While Luxembourg has a lot of open data, processing, conflating, importing and maintaining it in OSM can be very time consuming. The localities, streets and house numbers dataset released by the government would be a great candidate to enrich OSM. Unfortunately, this dataset isn't perfect, and can be a bit of a headache. It is not versioned, is prone to change does not have stable identifiers and does not always contain the street name as spelled on the street signs. While reporting issues to the data set provider has worked in the past, it is hard to coordinate and keep track of the change requests. In the official data set, addresses are linked to a cadastral parcel and will by default be placed on the centre of largest building on that parcel. This causes some parcels to contain many buildings and addresses, as is the case in Schifflange, with up to 50 nodes on one single parcel. When they haven't been manually adjusted in the official dataset, they all share the same position. When a parcel gets split, the address ends on both new parcels. This causes more than 2000 duplicated identical nodes. However, a local survey isn't always necessary to determine which building the address really belongs to. There are also some edge cases that require different processing, such as localities (81) without street names. These only contain a house number and a locality.

The current quality assurance process is also very time consuming and only supported by ad-hoc scripts, as some newer contributors manually copy data from the upstream data set, which introduces unprocessed data into OSM.

Mappers elsewhere have built tools that fit their local data and issues: https://berlinadd.morbz.de in Berlin, https://github.com/britiger/osm_address_db/ in Brandenburg, https://regio-osm.de in Germany, https://github.com/rubund/addrnodeimport in Norway or https://gitlab.com/OSM-DK/autoAWS/ in Denmark. These provide precious inspiration and can be mined for parts, but can't be reused directly. Over the years, a couple of scripts and queries have been written in Luxembourg that solve different aspects of this larger project, but they have yet to be combined into a larger tool.

The need for a tool-supported process to manage the addresses has been discussed previously within the community, but time constraints make it hard for this to finally happen. Using the microgrant funding, a developer can implemented an open source tool-based process that uses the data set to enrich OSM.

Budget breakdown

Give your budget breakdown here. If you would like to present your budget in a tabular format, you can use this tool to convert your spreadsheet into a wikitable and paste the code below here

Item What do you plan to spend the money on? Expected cost per unit (in Euros) Quantity you plan to buy / subscription length Expected total cost (in Euros) Why is it needed? (Please give as much detail as possible.)
Working day (8h) Development 200 EUR 9 1800 EUR Described in "What do you need the grant money for?" and "If you receive a grant, what do you aim to achieve?"

If you receive a grant, what do you aim to achieve?

It is recommended that this supports mission of the OSM Foundation and your local community. Please provide us with any information you think is relevant including the metrics you expect to achieve. Please write 100-200 words.

The main goal is to improve the overall address information in Luxembourg, which will make OSM the best source for addresses in Luxembourg! This will also allow contributors to improve other aspects of OSM once the addresses are complete. In the long term, the result of this project is to unify the spelling of street names on OSM, in the official database and on the street signs on a national level.

The goal can be decomposed into the following tasks:

  • Improve the tooling used to process the official data source
  • Identify addresses that appear multiple times in the official dataset (typically happens when a cadastral parcel gets split) and multiple addresses that are clumped together (happens when a cadastral parcel contains more than one address). Create a map showing them and whether they still need to be surveyed for OSM.
  • Compare the official dataset and OSM and identify addresses that differ significantly. Provide a simple way of marking the official dataset as wrong.
  • For addresses where no street with the same name can be found, suggest possible corrections to addresses and streets to mappers, considering possible similar nearby street names.
  • Build lists and maps of addresses missing in OSM.
  • Identify addresses that have been newly created or modified in the official dataset. Create lists of differences for mappers to periodically review.
  • Build lists and maps of addresses missing or misplaced in the official dataset.

Do you receive funding from any other sources at the moment?

If you receive funding from other organisations currently, please state how much this is, whether the donations are regular or one off, and how these funds have been raised. Please write no more than 300 words. If you have any sources of co-funding for this grant application, please provide details. Examples of this might be if you have any partners who would be willing to match grant funding if you receive it, or if your project has existing funding. Please provide as much detail as possible.

No.

Is there anything you would like OSMF to support you with?

If there is anything which you would like OSMF to support you with in order to strengthen your project, please let us know. For example, this could include building relationships with OSM communities, sourcing updated imagery for mapping projects, or guidance on how to use some of the tools we have created. We want to do our best to support OSM communities, so please let us know what you need. Please write no more than 150 words.

It would be nice to be able to send a pre-approved message to all contributors within a given region. The home location, if set, could serve as the matching criteria.

Declaration

By submitting this form to the OpenStreetMap Foundation, you certify the information contained in this application is correct, and that if you are awarded a grant, you will use it only for the purposes described above. You will provide written documentation and receipts for all of your expenses to the OSM Foundation to demonstrate this. You understand that the decisions made by the OSM Foundation Microgrants committee are final.

David Morais Ferreira (dmlu)

Endorsements

Community members are encouraged to endorse your project request here!

  • Happy to endorse this project, this is exactly the sort of thing we need to make OSM a fully viable alternative to commercial geo providers - Ed Freyfogle