Of Science And Industry Map

Getting Started With Of Science And Industry Map in QGIS

The Science and Industry Map is a specific OpenStreetMap-derived tile style that renders its basemap with a different aesthetic and labeling hierarchy than the standard OSM layer. The tile URL is https://tile.nextzen.org/scienceindustry/19/{z}/{x}/{y}.png. You don't need anything fancy to use it. Add it as a XYZ tiles layer in QGIS and you are done. This style is actually useful for field documentation and printed materials because the industrial and scientific POI labels are given more visual weight than on the default basemap. The standard OSM tile layer is cluttered for this purpose. Research stations, factories, and testing facilities tend to stand out better here. The tradeoff is that roads and residential labels drop down the priority ladder faster. If you are building a web map rather than a static print layout, this layer works fine as a fallback basemap. For interactive use, load it alongside another basemap so your users can toggle between them.

Adding the layer in QGIS

Open QGIS and go to Layer menu, then Add Layer, then Add XYZ Tile Layer. Click New and enter a name like Science and Industry Map. For the URL template, paste: https://tile.nextzen.org/scienceindustry/19/{z}/{x}/{y}.png Click OK, then OK again. The layer will appear in your Layers panel. Zoom to your area of interest and it should render within a few seconds depending on your internet connection speed. If it does not load, check that you do not have a corporate proxy blocking tile access, which is a fairly common issue.

Exporting a usable tile set

Sometimes you need these tiles offline. Download the entire tile range for your study area using a tool like mbtiles-export or a browser extension. Cover the full zoom range from 10 to 19. A typical European country at that range produces roughly 4 to 6 GB of data, depending on how much of it was already cached in your browser. Load the mbtiles file back into QGIS via the MBTiles connection interface instead of the XYZ Tile interface. Performance is noticeably better because QGIS reads the local SQLite file directly rather than hitting a remote server on every zoom change.

A specific problem I ran into

I was building a map of research facilities around the Rhine valley, and the Science and Industry layer completely omitted several large water treatment and analysis labs that were clearly tagged in OSM as industrial facilities with additional amenity qualifiers. The standard OSM basemap rendered those same features without issue. The workaround was to enable both basemaps, compare the two side by side, and use the standard layer as a catch-all for features the Science and Industry style dropped. I then exported the combined data as a vector layer using the Overpass API, filtering for nodes and ways with tags like amenity=research_institute, amenity=university, industrial=yes, and man_made=works. That gave me complete coverage regardless of what the basemap rendering engine chose to show.

Common pitfalls

Do not assume this layer is a complete representation of every science or industry feature in OpenStreetMap. Tile styles are rendering configurations, not data sources. The underlying OSM data is the same. What changes is what gets drawn at each zoom level and how labels are prioritized. Another issue is label overlap. At zoom 15 and above, the Science and Industry style tends to place facility names in positions that can obscure nearby roads. This is mostly cosmetic, but it matters if you are producing a map where road network readability is important.

When this approach fails

If your project requires displaying real-time data updates on top of a basemap, this layer is not a good fit. The tile cache is hosted by NextZen and they do not offer any real-time update guarantee. If a new research station opens tomorrow, the tiles may show it within days, but there is no SLA for that. For time-sensitive applications, use a vector tile basemap instead, where you can refresh the data source independently. The Science and Industry tile layer is slower to load than the standard OSM basemap in many cases. That is because the rendering includes more icon-heavy POI symbols for industrial and scientific features. On slower connections, you may notice blank tiles popping in gradually rather than a smooth fade-in effect. The standard OSM layer handles this better because the rendering pipeline is simpler. If you need a lighter alternative, try the OSM Bright or CartoDB Positron basemaps. They render faster and provide a cleaner canvas for adding your own data on top. Use the Science and Industry layer only when the specific POI labeling style is actually needed for your final output.

Building a production map layer from this data

For a permanent solution, query the Overpass API directly for science and industry features in your area, save the results as a GeoJSON or PostGIS table, and style it yourself inside QGIS. This removes the dependency on any external tile server and gives you full control over which features appear at which zoom levels. The query is straightforward. Target nodes, ways, and relations with amenity tags that include research_institute, university, school, hospital, laboratory, and commercial tags for industrial_zone, factory, plant, and works. The query takes roughly 3 to 8 seconds to execute for a medium-sized region, depending on how much data matches. Style the resulting layer with point symbols that differentiate research institutions from industrial facilities. A blue circle for science, an orange square for industry. Keep the symbol size under 6 pixels so the map stays readable at small zoom levels. Export the project as a QGIS project file and share it. The entire workflow from query to styled map takes about 20 minutes on a normal machine.

This method is more work upfront than just dragging a basemap into QGIS, but it pays off quickly if you need the map to look consistent across devices and internet conditions. Tile servers can change, break, or become unreachable without warning. Your own data never disappears.