Examples - Wherobots

Choose an example by the result you want to produce. Each page includes the notebook code and instructions for running it in Wherobots Cloud.

Start here

Work through these four notebooks in order to learn the core workflow.

Load and query spatial data — Combine NYC buildings with elevation data and map the results.
Read spatial files — Load GeoParquet, GeoJSON, CSV, and raster data.
Prepare data for faster queries — Create managed tables and export GeoParquet with spatial metadata.
Join datasets by location — Count places within boundaries and find nearest neighbors.

Outcome	Notebook
Count California parcels and property value exposure by FEMA flood zone and city.	Coastal flood exposure
Read spatial data and call geometry functions in Scala.	Spatial analysis in Scala
Turn Overture buildings and roads into PMTiles for a map.	Vector tile generation
Filter TIGER railroads to Texas and generate a PMTiles archive.	Railroad PMTiles

Outcome	Notebook
Load GeoParquet, GeoJSON, Shapefiles, and GeoTIFF from S3.	Spatial file formats
Load imagery catalog items from a STAC endpoint into a DataFrame.	STAC reader
Query and map Overture buildings, places, and transportation data.	Overture Maps
Filter NOAA hailstorm observations by date and location, then map them.	NOAA storm data
Read a Unity Catalog Delta table, add a distance feature, and write the results back.	Unity Catalog ETL

RasterFlow examples cover imagery preparation and model inference. See RasterFlow requirements before running them.

Outcome	Notebook
Build seasonal or custom-date Sentinel-2 mosaics.	Sentinel-2 mosaics
Build a mosaic from NAIP aerial imagery.	NAIP mosaics
Discover your raster inputs with STAC and build a mosaic from a tile index.	Bring your own rasters
Estimate canopy height from aerial imagery.	Canopy height
Detect roads in aerial imagery and vectorize the predictions.	Road detection
Compare imagery from two dates to detect changes.	Change detection
Detect crop fields in Sentinel-2 imagery and save field polygons.	Field boundaries
Detect sidewalks in aerial imagery and create network geometries.	Sidewalk detection
Detect objects in aerial imagery using a text prompt.	Text-prompted detection
Export a PyTorch model and run it through RasterFlow.	Bring your own model
Load ESA WorldCover tiles and visualize land cover.	ESA WorldCover
Summarize land cover around Overture building footprints in Texas.	Zonal statistics

Outcome	Notebook
Group nearby spatial points into density-based clusters.	DBSCAN clustering
Identify spatial hotspots with Getis-Ord Gi*.	Hotspot detection
Score spatial outliers using Local Outlier Factor.	Outlier detection
Join each feature to its K nearest neighbors.	Nearest-neighbor joins
Match vehicle GPS trajectories to OpenStreetMap roads.	GPS map matching
Calculate the area reachable within a driving time.	Drive-time isochrones

For examples organized by a business question, browse industry use cases.