OpenStreetMap Blogs

I have to mapping projects I’m doing here. The first one, the one that has most of my focus is my home community of Fairview, California. The other is Canonsburg, Pennsylvania. For both projects I rely on a variety of tools:

• the Bing visual
• Google maps
• County GIS records

While Google has good data most of the time, I have spotted a couple dozen errors since I’ve started on these projects. When I can easily submit a correction to them, I do. When I can’t, I make sure my work better reflects what the county map tells me.

04/12/2025-10/12/2025

Mapping campaigns

• confusedbuffalo is proposing to extend their existing automated edit of phone numbers in the USA and UK to include Canada, removing extra punctuation and adding the country code. Mappers are invited to discuss this on the OSM Community forum and to get in touch if they think such an edit would also be useful and accepted in their country.

Community

• Raquel Dezidério Souto has published an entry in her OSM user diary about the tagging scheme developed for the collaborative mapping of beach accesses in OpenStreetMap. The researchers hope it can be useful to show how versatile OSM is, facilitating detailed mapping of these access points, as well as drawing attention to the value of beaches as public areas that should remain accessible to all.
• David Alberto has made a map that visualises the distribution of sundials in Europe using OpenStreetMap data (amenity=clock + display=sundial).
• Hobbesvsboyle gave an introductory presentation about OpenStreetMap to an audience of local bike advocates in Madison, Wisconsin. He has made the slides, notes, and a resource sheet available under a Creative Commons BY-SA license so that others can re-use it.
• Oldnab recounted several of his encounters during his mapping walks.
• Matteo Schizzerotto has transformed a car dashcam into a semi-automated OpenStreetMap pipeline complete with GPX extraction, a React/FastAPI helper app, data-based clip prioritisation, and Valhalla routing tweaks. This slashes manual work, highlights the most valuable footage, and directs surveys towards new roads.

OpenStreetMap Foundation

• The OSMF State of the Map Organising Committee has announced that the 2026 State of the Map will be held in Paris 28 to 30 August 2026. Visit 2026.stateofthemap.org for updates on speakers, the agenda, ticketing, and ways to get involved as the event approaches.
• Enock Seth Nyamador reported that twenty-five grantees of the State of the Map 2025 Travel Grant Programme were able to attend the event in Manila.
• The SotM 2026 Organising Committee is asking the community for a logo for the next SotM, to be held in Paris from 28 to 30 August 2026. Logo proposals should be submitted before 15 January 2026 23:59 UTC.

Local chapter news

• OSM Latam has opened ► submissions for the 2026 State of the Map Latam logo competition. Entries will be accepted until 9 January 2026, after which they will undergo community voting and a committee review. The winning design is scheduled to be revealed on 28 January 2026.

Education

• The Railways of the World YouTube channel has demonstrated how a navigation app works by using graph theory, OpenStreetMap data, and OSRM, in a video titled ‘How I Beat Google Maps With My Own Navigation App’.
• Paul has published, on AlternativeTo, an article titled ‘From quest‑hunting to cartographer‑craft: a guide to contributing to OpenStreetMap’, with an overview of popular software used with OSM.

Maps

• Osmldf announced ► that Architectes dans OSM, an initiative to enhance OpenStreetMap building data by adding their architects and construction dates, has transitioned to its own dedicated platform at mapaedifi.eu. There is also a uMap to visualise these features.
• dktue has made a plot of OSM data density and it is available in Wikimedia Commons, under a CC0 license.

OSM in action

• [1] Gadjah Mada University has developed the ‘Geoportal Informasi Kebencanaan’, an OpenStreetMap-based web map that allows survivors of the recent floods and landslides in Sumatra to map the disaster’s impact and pinpoint village locations requiring urgent logistical assistance.

Open Data

• Mila Janovoso wrote about GlobalBuildingAtlas, an open global and complete dataset of building polygons, heights and LoD1 3D models. You can read the paper by Xiao Xiang Zhu (Technical University of Munich) and others published in Earth System Science Data.

Software

• An architecture student has created ArchiKEK to convert OpenStreetMap data into layered, design-ready SVGs for site analysis, offering one free high-resolution export with a style tailored for architectural use.
• From December 2025, Magic Earth will switch ► to a premium model for its navigation app, whereby key features such as offline maps and real-time traffic will only be available with a EUR 14.99/year subscription, while basic navigation will remain free of charge.
• OSM Wrapped is a cool platform that shows and celebrates key overall and annual contributions of OpenStreetMap members. Data is fetched from the HDYC Platform and OSM heatmap. Credits to Kennedy Kambi, Joseph Munienzy, Pascal Neis, and OpenStreetMap contributors.

Programming

• Astrid (aka aselnigu) has published in her diary about the lack of some road segments in OSM depending on the zoom level, as commented on the OpenStreetMap Community forum. The only difference between the two road segments was that the visible part contained one additional tag (maxspeed=*) and she did some trials in the Lua language to understand what was happening.

Releases

• Christoph Hormann has released an updated Styleinfo analysis for both the Alternative Colours style and the OSM-Carto style, offering an overview of how each rendering approach presents OpenStreetMap data.

OSM in the media

• Johnny Harris speedran the Swiss rail system, using OpenHistoricalMap to explain Swiss history (segment from 4:30).

Other “geo” things

• IGN has built a beta for visualising CityGML LoD2 rendering of buildings based on their latest LiDAR survey.
• Miguel García Álvarez (aka Milhaud) has published ►, in the blog Mapas Milhaud, about a map that shows the mythical beasts of Germany, created in 2024 by Neil Parkinson, from Púca Printhouse. There are also mythical beasts maps of the Netherlands, Italy, Wales, Scotland, Iberia, England, France, Ireland, Canada, and more.
• Stefan Krempl, of Heise, reported that researchers have observed increasingly complex and intensified GNSS interference in the Baltic region, raising significant concerns about the operational safety of cargo and passenger vessels. The interference becomes markedly stronger in open waters, suggesting the possible involvement of a high-powered military system.
• Stadia Maps enables developers to add real-time and predictive traffic data to their routing API, giving users accurate ETAs based on live conditions and historical trends without changing their integration or sacrificing privacy.
• Bart Holterman et al. have developed Viabundus, an online map of long-distance routes in late medieval and early modern Europe (1350–1650).

Upcoming Events

Country	Where	Venue	What	When
		Online	OpenStreetMap Midwest Meetup	2025-12-11
			OSMF Engineering Working Group meeting	2025-12-12
	Berlin	Lucia Weihnachtsmarkt	210. OSM-Stammtisch Berlin-Brandenburg	2025-12-12
	Zürich	Bitwäscherei Zürich	182. OSM-Stammtisch Zürich	2025-12-12
	Zaragoza	Etopia	GeoCamp – State of the Map España 2025	2025-12-13
		MAP Mercator museum	OpenStreetMap Belgium at the MAP-Mercator museum	2025-12-13
	Jalpaiguri	Thana More	10th OpenStreetMap West Bengal Mapping Party	2025-12-13
	Navi Mumbai	Smt. Indira Gandhi College of Engineering (SIGCE)	Outdoor Mapping Party @ MiniDebConf, Mumbai	2025-12-14
	København	Cafe Bevar’s	OSMmapperCPH	2025-12-14
	Hannover	Kuriosum	OSM-Stammtisch Hannover	2025-12-15
	Budapest	Cartographia Kft.	OSM térképest – december	2025-12-16
	Madrid	Online	Mappy Hour OSM España	2025-12-16
			Missing Maps London: (Online) Mid-Month Mapathon [eng]	2025-12-16
	Lyon	Tubà	Réunion du groupe local de Lyon	2025-12-16
	Bonn	Dotty’s	195. OSM-Stammtisch Bonn	2025-12-16
	City of Westminster	Monkey Puzzle pub	London xmas pub meet-up	2025-12-16
		Online	Lüneburger Mappertreffen	2025-12-16
			Réunion du groupe local de Vienne	2025-12-17
	Düsseldorf	Online bei https://meet.jit.si/OSM-DUS-2025	Düsseldorfer OpenStreetMap-Treffen (online)	2025-12-17
	Karlsruhe	Chiang Mai	Stammtisch Karlsruhe	2025-12-17
	Stainach-Pürgg	Online	19. Österreichischer OSM-Stammtisch (online)	2025-12-18
			UN Mappers Mappy Hour	2025-12-19
	Potsdam	Luisenplatz, Potsdamer Weihnachtsmarkt	Potsdamer Mappertreffen	2025-12-19
	Kalyani Nagar	TomTom, Shanti One, Ramwadi, Pune	Indoor Mapping Party @ TomTom Pune	2025-12-20
		MAP Mercator museum	OpenStreetMap Belgium at the MAP-Mercator museum	2025-12-20
			Missing Maps : Mapathon en ligne – CartONG [fr]	2025-12-22
		MAP Mercator museum	OpenStreetMap Belgium at the MAP-Mercator museum	2025-12-27
	New Delhi	online	OSM India×TomTom Online Mapathon	2025-12-28

Note:
If you like to see your event here, please put it into the OSM calendar. Only data which is there, will appear in weeklyOSM.

This weeklyOSM was produced by MatthiasMatthias, Raquel Dezidério Souto, Robot8A, Andrew Davidson, barefootstache, derFred, mcliquid, s8321414.
We welcome link suggestions for the next issue via this form and look forward to your contributions.

The award results of the 2025 OSM EWG Microgrant Program are in!

 WaterwayMap.org 

WaterwayMap.org displays how waterways in OpenStreetMap (OSM) are connected, helping mappers find and fix tagging mistakes. The site maintains a river database based entirely on OSM data and is used as a reference source by entities like Wikipedia.
The grant covers approximately one year of web hosting costs for this service. (Read more.)

Map Review Team

The Map Review Team project supports teams of mappers in collaboratively reviewing changes to the map within a specific geographical area. Its grant will support integrating with existing OSM tools and APIs, such as WhoDidIt and OSM Notes, to gather event data. A key feature is the introduction of an “approval” status, allowing teams to focus their efforts on changes that have not yet been reviewed by another team member. (Read more.)

Field Tasking Manager

HOT (Humanitarian OpenStreetMap Team) will use their grant to enhance their Field Tasking Manager (FieldTM) tool. Their development focuses on improving OSM integration by creating compatible survey forms and developing a robust conflation workflow to merge field-verified data with existing OSM data. The plan includes partnering with a local OSM community/university chapter in a priority country to test the tool and field-verify a city’s data. (Read more)

UseOSM

Developed by the Unpatterned Lab team, UseOSM hels discover unused opportunities in OSM data. Its platform is live at useosm.org but is not fully operational. The launch is projected to be in January 2026. (Read more)

OSM Apps Catalog

The OSM Apps Catalog helps people discover apps that use OpenStreetMap. The grant will be used to redesign the landing page to look more like an app store, including: What’s new, daily featured apps; Language support; new categories and improved search; and a support/donate (to apps accessed) function. View the current platform at osm-apps.org. (Read more.)

StreetComplete

The StreetComplete team will use the grant to help encourage users to be more involved in the OSM community. They are working on ways to notify about nearby community events, list local community channels, and integrate Weekly OSM news. (Read more)

OpenStreetMap Stats Generator

OpenStreetMap Stats Generator reports stats about user contributions in mapping campaigns. They will use the grant to revamp the project after some years of not being maintained. The’ll move from Planet files to Parquet and are targeting SotM 2026 for the working version release. (Read More)

A few kilometers from the town of Luncavița in Tulcea County, tourists visiting the Măcin Mountains National Park can stop off in Cetățuia hamlet — a paradise for beekeepers who have brought bees to the linden forests (that define this hamlet) and for tourists who want to stay in a quiet place, next to the mountains, as this is the only locality in Tulcea County that resembles a mountain village. However, it didn’t have detailed OpenStreetMap coverage, until now. We have mapped this area in detail and are pleased to announce that the map is complete. You can view the before-and-after images below.

Before detailed mapping

After detailed mapping

Notice: This changeset was made on the 29th of november (15 days ago as of writing this diary.)

নাগা বাজার — বীরকুৎসা সড়ক আমাদের দেখায় কীভাবে একটি ছোট স্থানীয় পথ থেকে বৃহৎ পরিবর্তন ও উন্নয়নের সম্ভাবনা তৈরি হতে পারে। এটি কেবল পথ নয়—এখানে মানুষের জীবিকা, সংস্কৃতি, অর্থনীতি ও দৈনন্দিন জীবনের রক্ষাকবচ লুকিয়ে আছে। সঠিক পরিকল্পনা, স্থায়ী রক্ষণাবেক্ষণ, স্থানীয় অংশগ্রহণ ও পরিবেশের সঙ্গে সমন্বয় রেখে উন্নয়ন করলে এই ৩ কিলোমিটার পথ পুরো এলাকাকে নতুনভাবে সংজ্ঞায়িত করতে পারে। রাষ্ট্রীয় তহবিল, স্থানীয় সরকার ও গৃহস্থালি উদ্যোগ মিলিয়ে এই সড়কটিকে শুধু চলার পথ না রেখে একটি স্থায়ী উন্নয়নের কাঁচা-সহজ রিংক হিসেবে গড়ে তোলা যেতে পারে—যা আগামী প্রজন্মের জন্য উন্নত সুযোগ ও জীবিকা তৈরি করবে।

Orange is pleasant to walk around. The houses are unique and interesting, the people are friendly, and sometimes there are dogs outside you can say hi to (and I say hi to every single one). Even in the quieter neighborhoods, it’s still a lively place. At least, during the day.

At night, it’s perfectly quiet and still. Because of this, I like to walk around at night; sometimes to help digest a large meal, sometimes just to think a bit before bed. Despite the solitude, I don’t feel entirely alone, because I’m being watched by a network of doorbell cameras. Some of them are set far enough back that I’m pretty sure they can’t see me. Others are close enough that they come alive and emit a faint glow when I walk by. Can they see in the dark? Does someone get a push notification alerting them to my presence? I’m not breaking the law, but it feels like I am.

I’m interested in privacy rights, and how technology makes it easy to surveil entire groups of people, always without their consent. I believe that people have a right to know when they’re being watched, and by whom, at least as long as recording in public remains legal. To that end, I’m starting to map doorbell cameras.

I did some research beforehand to see what the community thinks of mapping private objects. I’m still new, and I want to make sure I contribute in a way that is welcome and helpful. The Good practice page also reads: “Be brave in what you add…”

I read the Mapping private information wiki page. As far as I can tell, nothing there explicitly prohibits doorbell camera mapping. I believe I can also satisfy verifiability requirements: someone with a reasonably up-to-date prescription could stand on the same sidewalk and verify that these devices exist at a given location.

I also found a forum thread with some lively debate. The arguments I read in favor were convincing. Then, I checked the map, which had 1,121 camera:mount=doorbell tags, some of which go quite far back.

Solar panels and swimming pools, also private objects, are mapped from imagery all the time. It’s not difficult at all to see several of the cameras I mapped on Google StreetView, so this isn’t even a theoretical thing. Whether or not you own a doorbell camera, foliage and delivery trucks permitting, has already been made public by Google anyway. (Ironically, one of the cameras I mapped was for a house whose owners have had their house removed from StreetView.)

I would also make a moral argument for mapping them. These are private devices which nevertheless do sometimes record public spaces, and then send that data to corporations that don’t have our best interests at heart. I believe everyone has a right to know exactly when, where, and how they’re being recorded. (I would rather people not be recorded at all, but you work with the laws you have, not the laws you want.)

If we will continue to not have a reasonable expectation of privacy in public, the next best thing we can do is inform ourselves and others of when they’re being surveilled by corporate interests, especially people who have no choice but to interact with surveilled residences. Delivery drivers and mail carriers don’t get a choice about whether or not they’re recorded by these devices.

So, after my walk a couple nights ago, I decided to add some of this data to OpenStreetMap. With just what I could see from the sidewalk–I want to make it very clear that I never once set foot on private property–I added nine doorbell cameras from a survey. I have already added more.

I was partially inspired by the recent news coverage of Flock, the ALPR surveillance company (with many cameras in Southern California) that provides a service that law enforcement can use entirely warrantlessly. Some law enforcement officials have already used this information to stalk or harass people they know.

If nothing else, seeing our neighborhoods lit up with doorbell cameras like so many Christmas trees will tell us a lot about the places we live.

Peta Bencana Banjir Kabupaten Aceh Tamiang

gpx desa terdampak
PDF Ukuran Kertas A-0
Google view

Current wiki guidance for tagging slipways (leisure=slipway) is as follows:

A typical mapping might be to make a short highway=service way to the water’s edge and make the final node a leisure=slipway.

The highway=service that is part of the physical slipway is often further refined with a service=slipway. Keep in mind though that this does not replace leisure=slipway, mapped as a node on the end of that way.

Some mappers instead tag the whole way as leisure=slipway without using a highway=* tag, or sometimes in combination with highway=service (here service=slipway can be applied as well as above).

In other words there are two options (the first being implicitly preferred)

Option 1 - tag the node

• way: highway=service + service=slipway
• final node: leisure=slipway

Option 2 - tag the way

• way: leisure=slipway + optional highway=service + service=slipway
• final node: not tagged

The map marks an icon for both nodes and ways tagged with leisure=slipway. It’s therefore likely not advisable to do both. But which one is preferable?

There is some discussion about this on the wiki but no apparent resolution and it appears to refer to previous versions of the guidance: osm.wiki/Talk:Tag:leisure%3Dslipway

Though there is a question raised about wether leisure=slipway should be replaced by highway=service + service=slipway outright, especially given not all slips are for leisure use. But the map only marks an icon for leisure=slipway so this is likely not recommended.

Status quo

I wrote a script to analyse the data from of an area on the south east coast of Scotland.

https://overpass-turbo.eu/s/2hdk

The results:

• 45: option 1 (includes 16 orphaned nodes with no way)
• 62: option 2

So option 2 appears to be preferred, however it results in the map icon being placed at the mid point of the slip.

For ways that are correctly scoped to just the sloped slipway, this is probably ok. Longer ways should be split anyway.

But tagging the final node could be a more useful map display in some cases.

What do other mapping services do? I don’t have large enough scale Ordnance Survey maps or otherwise to hand to be able to tell properly. My Imray chart seems to place the slip icon at the end of the slip but the scale is relatively small.

The one thing that feels clear is that we should only use leisure=slipway on either the way or final node, not both. Also, ways should be properly scoped and nodes should have an associated way.

Details

Analysis script: https://gist.github.com/jwheare/a3577de092c68b1789e454e832b38f5b

Overpass query:

https://overpass-turbo.eu/s/2hdk

[out:json][timeout:90][bbox:55.37911,-4.21875,56.944974,-1.40625];
(
  // nodes only
  node["leisure"="slipway"];
  // ways that contain the above nodes
  way(bn)["highway"];
  // all nodes/ways/relations
  nwr["leisure"="slipway"];
  nwr["service"="slipway"];
);
out geom;

Appendix

Node not tagged (62)

• 35: leisure=slipway highway=service service=slipway ✅ perfect option 2
• 25: leisure=slipway
  • could add highway=service + service=slipway
• 2: service=slipway highway=service
  • doesn’t use leisure=slipway at all, should be fixed (done)
    • way/1012902378
    • way/1346501677

End node tagged (29)

• 15: highway=service
  • could add service=slipway
• 9: highway=service service=slipway ✅ perfect option 1
• 3: leisure=slipway highway=service service=slipway
  • double tagged, should be fixed (done)
• 1: highway=footway
  • way/57775872
  • long footway that ends in a slip. should be split up into two separate ways (done)
• 1: highway=service service=driveway
  • way/129911891
    • short driveway that ends in a slip. should be split up into two separate ways (done)

Start node tagged (8) (way should be reversed - done)

• 6: highway=service
  • could add service=slipway (done)
• 1: leisure=slipway highway=service service=slipway
  • double tagged, should be fixed, but complicated arrangement so unclear which option is best
    • way/210856759
• 1: highway=path
  • long path that ends in a slip, with the start node. should be split up into two separate ways with the slip reversed (done)
    • way/978785415

Some other node tagged (4)

• 2: highway=service
  • way/327317264
    • long track that ends in a slip, with the node tagged in the middle. should be split up into two separate ways (done)
  • way/626239129
    • doesn’t need to be split, but node should be moved (or removed and way retagged) (done)
• 2: highway=track
  • way/887901902
    • mistagged car park with a slip at a point where it meets water. should have a separate way added for the slip (done)
  • way/499488192
    • long track that ends in a slip, with the node tagged at the top. should be split up into two separate ways (done)

Stats

• nodes: 56
  • orphans: 16 (way should be added)
• ways: 103
  • leisure=slip node missing: 62
  • leisure=slip node at start: 8
  • leisure=slip node in middle: 4
  • leisure=slip node at end: 29
• way tags
  • highway=service (74)
  • leisure=slipway (64)
  • service=slipway (50)
  • highway=track (2)
  • highway=footway (1)
  • service=driveway (1)
  • highway=path (1)

Changesets made to fix various issues mentioned here:

• changeset/175892606
• changeset/175893398
• changeset/175893662
• changeset/175894013
• changeset/175894979
• changeset/175895389
• changeset/175896209
• changeset/175897032

To edit an OpenStreetMap (OSM) diary entry, you must navigate to the specific post while logged into your account.

Steps to Edit Your OSM Diary

• Log in to your account on OpenStreetMap.org.
• Click on your username in the top right corner to go to your user profile.
• Click the “New Diary Entry” link (usually located in the menu tab just below your username).
• Find the specific entry you want to change and click on its title to open the full post.
• Scroll to the bottom of the post text; you will see an “Edit this entry” button.
• Click it to open the editor, make your changes, and hit Save.

Important Notes

1. Deletion: There is no direct “Delete” button for diary entries. If you want to remove a post, the standard practice is to edit the entry, delete all the text (replacing it with something like “Post deleted”), and save it.
2. Formatting: The diary uses Markdown for formatting. You cannot use rich text (like in Word), so you will need to use specific characters for bolding, links, and images.
3. Images: You cannot upload images directly to the diary. You must upload them to an external site (or the OSM Wiki) first and then link to them using Markdown:

Buste op het Kennemerpark in Alkmaar.
Het OSM knooppunt
Meer informatie op de website van Kunstwacht.

Wijzigingset
Op OSM

wikidata:Q2369995
wikimedia_commons: Category:Dorpskerk (Castricum)
wikipedia nl:Dorpskerk (Castricum)

Rijksdienst voor het Cultureel Erfgoed, CC BY-SA 4.0, via Wikimedia Commons

Hoje, comecei uma ordem de trabalhos novos: fui à pagina da taginfo para key:building, ordenei os termos por quantidade de valores e fui à última página. Aqui estão nodes ou áreas etiquetadas como “buildings” em que só há uma instância de cada - ou seja, há 98% a certeza que está mal etiquetada. Lancei-me à correcção ordenada de cada uma.

É uma forma de ajudar a limpar as tags: de momento, 625 páginas de valores na chave “building”, 8120 itens… de certo agora 624 páginas pois editei os itens todos da última página e agora estou à espera que as edições cheguem aos servidores e actualizem esta tabela …e também de editar fora da minha por assim dizer jurisprudência, porque até agora cingi-me a mapear passeios, passadeiras, ciclovias, acrescentar vários detalhes, e corrigir problemas indicados pelo Osmose na minha área de residência.

Até agora, deparei-me com (entre outros!):

• uma área afectada por desastres naturais - neste caso, era um node, e tive de criar uma área nova e deixar lá o node com a informação dos estragos. Não era um edifício o que fez com que a pesquisa de como etiquetar este tipo de sítios fosse um bocadinho irrelevante para o caso, mas de qualquer modo aprendi que à falta de melhor, sempre existe damage=yes. Na área, pus landuse=farmland

• uma cabaninha Sufi no meio do nada. estava etiquetada como Teqe - em que resolvi um pedido de correcção de alguém que também reparou nesta tag solitária e gentilmente deixou esta página da wikipedia para alguém continuar a pesquisa. acabei por etiquetar como building=sufi_lodge, que também não está a ser usada por mais nenhum artigo no site… mas que está 100% correcta, porque é o que é.

• um bar dentro dum hotel na Bolívia, que já não era no hotel, mas que mudou para muito perto - e o desafio de encontrar que bar é que abriu no mesmo sítio - aqui posso ou não dar-me ao trabalho de acrescentar moradas, horários, e tal. Aqui só apaguei a tag de building, porque era um node dentro do hotel, e já especificado como amenity=pub, e mudei o nome do bar.

• um edifício que, de acordo com o resumo de edições anteriores, era escritório governamental e estava a ser usada agora como parte duma escola - não me pus a mexer muito, não ficasse ali trinta horas a mapear a escola toda. Deixei building=government_office e pus o nome das turmas que estavam a servir como chave nas notas

• um edifício em Inglaterra supostamente protegido como monumento histórico. não deu para perceber do que se tratava por isso não deu para confirmar. No caso só tenho de, meio resignadamente, mudar o valor para building=yes

• uma área conspícua no meio na Namíbia, que estava etiquetada como building… Pelo nome e com alguma pesquisa meia sofrida deu para perceber que era algo relacionado com um projecto de agricultura OU desbravamento para abrir estradas subsidiado pelo governo. Reparei que ao lado havia uma área mais pequenina que estava como landuse=farmyard, então só copiei. Deixei o nome do projecto e acrescentei o nome da área administrativa e o país.

Portantos… É uma ordem de trabalhos virtualmente infinita até se conseguir consolidar estes valores para algo mais aproximado da realidade. Sou capaz de ir indo fazer isto para outras chaves, aos bocados. Uma maneira muito específica de aumentar o conhecimento de tags no OSM e a cultura geral. Sei que há pessoal a criar grupos de trabalhos semi automatizados para tornar isto mais rápido (talvez?) e potencialmente mais divertido, mas não sei muito sobre isso ainda. Para já, vou desbravando páginas solitariamente! Será que já actualizou???

I was a bit too sleepy to post last night, but I got coffee and wandered by Long Dog Coffee and Treats, a delightful dachshund-themed café (with treats for both two-legged and four-legged friends) because that POI was not on the map. :)

Ahoj, tento projekt má za cíl zlepšit kvalitu a přesnost mapování obnovitelných zdrojů energie. A taky začít rozšiřovat infrastrukturu o tyto detailní data. Čím také přispěje k lepšímu pochopení a také budoucího vývoje v této oblasti energetiky.

Bod: 1. Opravy Mnohé např: fotovoltaické elektrárny jsou nesprávně zmapované jako například že se tváří že jsou vlastně solárním kolektorem a né fotovoltaickou elektrárnou.

Bod: 2. Přesné a detailní zpracování Je nutno dbát na kvalitu zapisování a mapování této infrastruktury.

Bod: 3. Expanze Růst této infrastruktury je klíčová k zlepšení kvality a dostupnosti dat o jejím vývoji. Je nutno přidávat nové generátory energie a elektrárny

https://accda.maps.arcgis.com/apps/webappviewer/index.html?id=2045711bf8a44a109d83d31300e5f5ed

Ashland, Cherryland, Hayward Acres and Fairview all used to have “Hayward” as their city for their addresses. The U.S. post office reverted their names to their historical place names. I am taking care of Fairview (I live here) but if anyone wants to help with the other communities, please jump in.

So as I was mapping some rooftop solar, I noticed that some of the houses were all offset from the Bing image data by the same amount. I’m pretty sure there’s a way to resolve this without turning my GPS on and driving around a bunch, but where’s the fun in that? I also wanted to play around with Go Map!! and get comfortable with uploading traces in general, so I turned on Go Map!!’s GPS and drove around a bunch while listening to a podcast. Seeing the GPS data later on in JOSM was neat! It was also a reminder that GPS has a sampling frequency, and moving in a car, even at 25 miles per hour, spreads those samples out.

I’m guessing this is an imagery offset problem (I think that’s usually more likely to be incorrect), but I want to be sure so the solar panels are correct. I have to dig in some more. I know there are various ways to do it, but I’m entirely too sleepy to figure them out right now, which means it is officially a tomorrow thing. :)

Edit: I got a brief second wind. I learned that there is an imagery offset database, although neither of the two options available helped in this instance. I also noticed that both Bing and ESRI imagery comport, so the buildings are offset by the same amount compared to both images.

I checked the history of one of the buildings, and it was created in 2018 from Bing imagery. I then checked the buildings against ESRI Clarity, and they match much better. Now it’s officially a tomorrow thing. ;)

I started mapping my community last month. Looking at how the map is filling in, I think I’m about 1/4 to a 1/3 done with getting Fairview on the map. I’m focusing on homes, property boundaries, trees, pools and other out buildings.

Tonight I also worked on sidewalks and crossings.

Was a map user with Mapinfo for years at work, and added Google Earth Pro to that skill set. AllTrails uses OSM for their base map. Recommended by Carey Spies.

Enjoying the learning process so far. Have edited the course cart path at the Grizzly. And have added the path/driveway to Dale Spies’ house.

Last weekend I gave a talk about OpenStreetMap and biking at the Madison Bike Advocacy Meetup. This is a monthly meeting of people interested in biking and local bike advocacy. The purpose of the talk was to give a high-level overview of what OpenStreetMap is, how it can be used for biking and bike advocacy, and how you can contribute to OpenStreetMap. About two-thirds of the audience had previously heard about OSM. I think the talk was well received and there were many questions during the Q&A period. And in the days after the event I saw some OSM editing activity from new contributors.

The slides, notes, and resources are available in this Github repository under a CC BY-SA license. Please feel free to reuse it!

All mistakes and omissions are mine.

• Suivi des commerces (@karr4s)
• Suivi des zones en travaux : requête à mettre à jour en fonction de la date + utilisation du tag check_date dans OSM pour avoir un suivi des travaux
• Suivi des vitesses limites

আজ আমি OpenStreetMap-এ নাগা বাজার ব্রিজের অবস্থান এবং নাম সংযোজন করেছি। নাগা বাজার ব্রিজ স্থানীয়ভাবে পরিচিত একটি গুরুত্বপূর্ণ সেতু, যা নাগা বাজার এবং আশেপাশের এলাকা সংযোগে ব্যবহৃত হয়। সেতুটি একটি প্রধান স্থানীয় রাস্তা পারাপারের জন্য ব্যবহৃত হয় এবং এলাকাবাসীর দৈনন্দিন জীবনে গুরুত্বপূর্ণ ভূমিকা রাখে।

আমি সঠিক অবস্থান নির্ধারণ করার জন্য স্থানীয় তথ্য ও GPS ব্যবহার করেছি। আশা করছি এই সংযোজন স্থানীয় মানচিত্রকে আরও তথ্যসমৃদ্ধ এবং ব্যবহারযোগ্য করবে।

Am 10. Dezember hat das Online-Vernetzungstreffen der FOSSGIS-Community stattgefunden. Mehr als 20 Teilnehmende nutzten die Gelegenheit, sich Wissen und Informationen zum Thema “KI in der Geodatenanalyse: Chancen und Herausforderungen” zu holen und sich auszutauschen und zu vernetzen.

Die Idee des Vernetzungstreffens, welches sich an Geoinformatiker:innen, FOSSGIS-Konferenz-Teilnehmende, Mitglieder und Interessierte in und an der FOSSGIS-Community richtet, ist, das gegenseitige Kennenlernen sowie Austausch zu Anliegen oder Themen, hier können sich Kontaktpunkte und Anknüpfungspunkte finden.

KI in der Geodatenanalyse: Chancen und Herausforderungen

Hendrik Wagenseil, Geograf, arbeitet beim BKG in der Arbeitsgruppe KI- Analyse von Fernerkundungsdaten. Nach ein paar Basics nimmt uns Hendrik mit in die Fragestellungen und Abläufe und erklärt anhand von Anwendungsfällen den Einsatz von KI beim BKG, beleuchtet Nutzen, Erwartung an KI, sowie auch die Herausforderungen. Als nützlich erweist sich die Ressourceneffizienz bei Prozessen, die viel manuelles Arbeiten erfordern, aber auch bei Änderungen in Daten kann KI unterstützen sowie es auch bei der Erschließung der Datenvielfalt von Lidardaten nützlich sein kann. Wenn ein Prozess operativ in den Einsatz kommt, ist die Schnelligkeit von Nutzen, weil damit viel händische Arbeit weg fällt.

Nutzen von Geo-KI

Noch ist das nicht soweit, die Verfahren befinden sich in verschiedenen Entwicklungsstadien und sind auf dem Weg zum Abschluss. Durch iteratives Arbeiten lassen sich die Prozesse Stück für Stück in den operativer Einsatz überführen. Es gibt Schnittstellen zu definieren, den Umgang mit Unsicherheiten zu üben und der Ressourcenhungrigkeit der Techologie gerecht zu werden.

Um mit der Schnelligkeit der aktuellen Entwicklung mitzuhalten, braucht es agiles und exploratives Arbeiten, hier machen es die Hürden einer Behörde einhergehend mit starken Sicherheitseinschränkungen und dazugehöriger Bürokratie, z.B. DSGVO nicht leicht.

Der Anwendungsfall Baustellenerkennung war beispielgebend für die Ausführungen im Vortrag und die Darstellung der “KI-Pipeline”.

Pipeline Geo-KI

Q&A

Die Teilnehmenden bringen diverse Hintergründe mit, beschäftigen sich mit OSS-Software, Open Data oder OSM im Job in der öffentlichen Verwaltung oder Privatwortschaft und haben verschiedenste Fragestellungen zum Thema.

• Frage: Wenn das Erkennungs-Modell auf einem anderen Modell aufbaut, kann man ihm auch “Vorkenntnisse” wieder abtrainieren? Z.B. Wenn viele Baustellen in den USA im Trainingsdatensatz waren und diese bestimmte Merkmale (z.B. Fahrzeuge) aufweisen die für die Baustellenerkennung hier irrelevant bzw. irreführend sind?
  Antwort: Architektur + Gewichte + Beispiele + Training mit False Positives macht das Modell genauer.

• Frage: Wie sieht es mit FOSS aus? Werden die Modelle open Source gestellt?
  Antwort: BKG hat die exotische Rolle in Bereich Datenharmonisierung, d.h. die Aufgabe ist es von den Ländern übergebene DOP-Daten zu vereinheitlichen und bereitzustellen. Grundlage ist eine Verwaltungsvereinbarung, die Daten gehören nicht dem BKG, deshalb können Labels nicht Open Source gestellt werden. Beispiel Daten zum Digitalen Zwilling gehören dem BKG, Veröffentlichung ist angedacht/geplant, da es sicherheitsrelevante Details enthält, gibt es dazu noch Diskussionen.
  Hier spielt auch das Thema Lizenzen rein, Sentinel-Daten sind frei verfügbar, Privatanbieter bieten kommerzielle Produkte, die teilweise detailreicher sind, eine Veröffentlichung ist deshalb nicht möglich bzw. unbezahlbar. Die zugrunde liegenden rechtlichen Regelungen geben verchiedenen Aussagen, das Haushaltsrecht regelt anders, als die Open Data Verordnung.
  Deshalb ist es noch offen, ob Code ververöffentlicht werden kann.
  Hinweis aus der Teilnehmerschaft: Falk Zscheiles Vortrag auf der FOSSGIS 2025 sagt, dass das Modell veröffentlicht werden dürfte.

• Frage: Gibt es Erfahrungen mit Digitalisierung von gescannten analogen Beabauungsplänen etc.?
  Antwort: Liegenschaftskataster ist Ländersache. Niedersachsen hat ein fortgeschrittenes Modell um einzelne Gebäude auf Luftbildern zu erkennen, hier sind Fernerkundungsdaten die Grundlage, keine B-Pläne.

• Frage: Hat das BKG Erfahrungen zum Georeferenzieren mit KI?
  Antwort: nein

• Frage: Headline: “Erstmals alle Gebäude der Welt als 3DModell verfügbar” - Ki oder Hexerei?
  Antwort: Mitteilung bezieht sich auf LOD1-Daten

Auch auf der FOSSGIS 2026 wird KI ein Thema sein, das Programm ist auf der Konferenzomepage verfügbar: https://fossgis-konferenz.de/2026/programm/.

FOSS4G 2025

Im nächsten Themenpunkt berichtete Mirko ein paar Highlights von der FOSS4G 2025 in Auckland, das ist die internationale FOSS-Konferenz im Geobereich an der ca. 500 Teilnehmende dabei waren. Das Thema Nachhaltigkeit wurde in vielen Facetten diskutiert, sowohl ökologisch als auch planerisch. Beiträge aus dem deutschsprachigen Raum wurden in diesem Jahr vermisst.

News aus dem Verein

Die FOSSGIS 2026 findet vom (24.)25.-28.03.2026 in Göttingen statt.

• das Programm ist veröffentlicht: https://fossgis-konferenz.de/2026/programm/
• Neu: lange Workshops am Dienstag, den 24.03.2026
• Anmeldung ab Januar 2026: https://fossgis-konferenz.de/2026/anmeldung/
• Sponsoren sind willkommen: https://fossgis-konferenz.de/2026/#Sponsoring
• Helfer:innen sind willkommen: https://fossgis-konferenz.de/2026/helfen/

Der FOSSGIS e.V. hat im Oktober einen Vereinsworkshop zur Weiterentwicklung mit 32 Mitgliedern veranstaltet, der Fokus lag auf verschiedenen Entwicklungsfeldern, wie Neumitglieder - Mitarbeit, bestehende Mitglieder - Unterstützung zur Mitarbeit, Nachwuchssicherung für Gremien, Welche Handlungsfelder sollten gestärkt werden?
Ein Ergebnis u.a. ist, das Vernetzungstreffen auszubauen, um die Vernetzung der Mitglieder untereinander zu verstärken.
Ein Bericht zum Vereinsworkshop ist im Newsblog veröffentlicht.
Die ersten Ideen zum Buddysystem wurden geschildert, ein Treffen zum Weiterdenken am Buddysystem ist für den 22.01.2026 um 15 Uhr vorgesehen.
Um an der Vision und Mission des Vereins zu arbeiten, gründet sich derzeit die AG Grundsatz, erste Infos sind im Wiki aufgeschrieben, das nächste Treffen findet am 19.01.2026 um 18 Uhr statt.

Der Vorstand des FOSSGIS-Vereins braucht neue Mitstreitende, eine komissarische Mitarbeit ist derzeit möglich, die nächste Vorstandswahl findet im Rahmen der Mitgliederversammlung im März statt. Fragen können an Katja oder jemand aus dem Vorstand gerichtet werden.

Zu guter Letzt gab es einen Aufruf zum Mitmachen an Aktivitäten des Vereins, beispielsweise bei der Unterstützung des FOSSGIS-Standes auf Veranstaltungen, wie re.publica 2026, Deutscher Kartographie Kongress (DKK) oder beim Open Source Park der OSBA auf der Smart Country Convention (SCCON), Maker Faire hannover, FrOSCon, Kielux oder KonGeoS. Bei Interesse und Fragen gerne an Katja wenden.

Die Aktivitäten des Vereins sind im FOSSGIS-Vereins-Kalender zu finden: https://fossgis.de/aktivitäten/termine.

Abschluss und Feedback

Zum Abschluss wurden in der Feedbackrunde das breite Themenspektrum als kurzweilig, ansprechend und abwechslungsreich benannt sowie die Kombination aus thematischen Informationen und Vernetzung den Verein erlebbar machen. Die Vorstellungs- und Vernetzungsrunde als interaktiver Teil der Veranstaltung wurde positiv bewertet, verbunden mit dem Wunsch nach mehr Zeit für Austausch dieser Art verbunden mit inspirierenden Fragen.

Das nächste Vernetzungstreffen ist für den 15.04.2026 um 18 Uhr geplant, die Vorbereitungsgruppe wird die entstandenen Ideen aufgreifen und freut sich auf das nächste Treffen.

Screenshot Vernetzungstreffen 10.12.2025

I’m finding it difficult to pick a particular mapping task to focus on, because even in my corner of Orange, there seems to be so much to do. For example, I’ve noticed that the sidewalk ways sort of peter out a few blocks away from the Circle. I was glancing at some geometry last night, and also noticed that some people have tagged a few backyard swimming pools, but there are plenty of unmapped pools visible on imagery. I spent a little time adding some pools, as well as adding access=private tags the existing ones. It’s nice to have some stuff to do that doesn’t always involve going out somewhere (not that that would be feasible in people’s backyards anyway heh).

Then I noticed solar panels on aerial imagery, and wondered if anyone had tagged them. I know people tag larger power infrastructure, but mapping rooftop solar seems like it would be cool data to have, too, especially in Southern California, where it works particularly well. It’s also easy to do from imagery.

The neighborhood with the pools also has a decent amount of rooftop solar, so I mapped a few of those as well. I realized, though, after uploading a bunch (oops), that some of the panels aren’t necessarily photovoltaic cells; they could be rooftop water heating, especially smaller areas that look particularly black. I plan to go back through and try to verify if they really are rooftop solar. Most probably are, and most of them look distinctive or big enough that I’d be confident in tagging them as rooftop photovoltaic cells, but a few of them could also be rooftop water heating, so I might just remove the photovoltaic-specific tags from them for now. It’s still rooftop solar, but the particular energy use will just be ambiguous. I think that’s better than it being incorrect.

Focus!

I need to pick a few things to focus on and them see them through to completion, I think. I have a bad habit of starting projects and not finishing them. In no particular order, I’d like to map:

• Fire hydrants (my OG quest)
• Rooftop solar
• Sidewalks
• POIs (especially adding business hours for things)
• Addresses for multi-family housing, although I’m not sure how OSM typically handles that, beyond interpolation.

The quixotic side of me would like to do this for the entire City of Orange, but that’s not feasible on my own. In the meantime, I’ll try to focus on a small area first and get that mapped as detailed as possible.

It would be nice to meet up with other mappers nearby, too. I’m not sure if OC has any mapper meetups, but I ought to check (and then maybe start one if it doesn’t!). If any OC OSM folks happen to stumble upon this, say hi!