Kendall County on OSM

Basemaps

How it all started

Esri Vector Tiles

Before

flowchart LR
L1[Centerlines]
L2[Hydro Features]
L3[Admin Boundaries]
L4[Etc.]

L1 & L2 & L3 & L4 --> Maps

After

flowchart TB
OSM -.-> E[Esri Tiles] --> Maps

Some years ago, Esri put out a vector tile basemap of OSM data. Before, we’d just been pulling our own internal layers into our maps for reference data and labels. This gave us the opportunity to get all those reference layers in a way that performed better, and we didn’t really have to think about the infrastructure. Win-win!

Editing OSM / New Mindset

• “If we’re gonna use this data, let’s make it awesome.”
• Stewardship, not gatekeeping

We started getting feedback about things like typos in road names, etc. How do we go about fixing those problems? Editing OSM! Just like that, editing the map became an official part of our department’s procedures.

It’s easy to see how making OSM better in our area directly benefits us, as data users, but also how it would beneft anyone using OSM. We couldn’t say the same about our internal layers.

Data Inventory

flowchart LR
1[Is it public?] --YES-->
2[Is it appropriate\nfor OSM?] --YES-->
3[Is there any reason to\nkeep an internal copy?] --NO-->
4[Is OSM as good\nor better?] --YES-->
5[DELETE]

3 --YES--> A[Figure out periodic\nedits / conflation] --> B
4 <--NO--> B[Edit OSM!]

At the end of this process, our “open data” portal has less in it, and is more specific to the things that the County is truly authoritative over. At the same time, more data is being added to and amended in OSM.

Current Uses / Projects

osm2pgsql

• Nightly update from GeoFabrik
• Basically everything
• All tags and metadata
• A bit of post-processing on route relations
• Centroids

Eventually we got tired of the update intervals and feature / scale limits imposed on the Esri tiles and wanted to do more with OSM data and have more control over it.

With a PostGIS database of OSM data, we can bring specific features into basically anything we want. We can also control the extent of the data so that we’re only dealing with our area.

Basemaps, again

One of the nice things about the PostGIS database is that basically anything can connect to it if you write the query correctly. Even though our Esri stuff needs to be published a certain type of way, we can bring the data into ArcGIS Pro with a Query layer, then publish it to our GIS portal that way.

Here we have highway label vector tiles, but with the kinds of shields I want, not a carto-lookalike.

Point of Interest Locator

Most folks on our maps are looking for addresses and parcels, and we already have that covered. But we thought it would be helpful to have a locator based on points of interest.

It ended up being an accidental tool for finding generic objects with a name tag, too.

But again, this shows easily we can bring OSM data into even an Esri environment.

Print Maps

Forest Preserves

Aside from the contours and surface texturing, this is 100% OSM data.

When someone asks us for a print map, one of the first things we do is map the heck out of it in OSM first.

Campus Maps

Again, 100% OSM data. Those aren’t tree points, those are individual shrubbery areas for the much around the trees. Highway areas, too.

Roads

Once upon a time, we managed the County’s emergency dispatch data. We also keep the data that feeds into the official highway maps. It’s not good, but we don’t have any say in the schema. We may never get other departments / agencies to let go of their internal layers and just use OSM, but since we use OSM, we can make the roads really high quality.

As much information as possible!

• operators (pictured)
• speed limits
• traffic signals and signs
• classification
• lanes
• surfaces
• areas

We want the operator tag on all the roads to clearly identify who’s in charge of a given street. We filled in everything we can, then sent maps to the local highway departments to ask for their help in verifying what we have and weighing in on roads that are unclear.

The project is ongoing, but we’re hoping to get a clearer sense of roadway jurisdictions through this process.

We’re gradually going through to update all the roads to reflect proper tagging for lanes, speed limits, and more. We don’t necessarily have a current project for these other details.

OSMUS Tasking Manager

Projects, Interns

• Buildings
• Sidewalks and Crossings
• Signs / poles
• Landuse
• Natural areas
• Manholes
• POIs
• Small hydro features

Buildings and sidewalks were picked for our crop of summer interns in 2023 to focus on. That was the focus of the herding cats talk.

The other projects are to help us organize our own work, and to have ready projects for future interns.

Maproulette

Centennial Farms

We were asked to map out “centennial farms” in the County. That’s farms that have been run by the same family continuously for over 100 years.

This is a great example of a project that involved data that was nowhere. What little information we had wasn’t even that good. But with the right historic tagging, it was totally the kind of thing that could be mapped in OSM. There are even signs for these places, so they could be verified by a survey.

We asked the state’s Department of Agriculture for the entire dataset, minus personal information, geocoded the result, and ended up with thousands of these farms. Not all of the resulting points were in sensible locations, though, so we put them on MapRoullette to be individually reviewed, and possibly added.

Restaurants POI Updating

Working with our Health Department, we help manage their Food Service Establishment database for inspections, etc. This dataset includes any establishment that prepares or sells food, so it is an incredibly rich dataset of restaurants, schools, grocery stores, etc, and one which is perpetually updated.

We’ve got permission from them to add non-sensitive information to OSM. The MR challenge isn’t live quite yet, but the goal will be to regularly conflate POIs in the county with this dataset, to ensure that these particular POIs are being kept up to date.

WikiData

Historic Plaques

Example: Historic Mill Stone Plaque

We were contacted by the County’s Planner to see if we could get some historic plaques added to the map. This information wasn’t in our system or anyone else’s that we could find.

We had the original newspaper article from the 70s that listed the plaques and their general location. Field trip!

Meagan drove around the county looking for the plaques, adding them to OSM as they were located and taking pictures. Once added, we made WikiData entries for each, and uploaded the images to WikiMedia Commons.

The Future!

CAMA Data

• Obvious stuff
  • Year built
  • Building Levels
  • Wall and roof materials
• Maybe? ¹
  • Room types / counts
  • Square footage

We have the “authoritative” layer of building points in the County! The quality of the township assessors’ work is at times questionable, but there’s often no other decent source for this data (except for real estate listings, but you often can’t copy that data).

1. This would really depend on some discussion and consensus with the community

Hypothetical!

Asset Inventories

• OSM Go + survey_date tag
• Routine check of quality, functionality
• Internal ref
• Join to internal asset DB on ref tag?

Questions?

Be in Touch!

josh@jcarlson.page

https://jcarlson.page

@jcarlson@mapstodon.space