Case Study: Geostellar achieves automated, high-resolution rooftop solar mapping and accelerated market expansion with Trimble eCognition

Using eCognition to Detect Solar Potential

Geostellar is an online solar marketplace that set out to make on-demand, e-commerce solar assessments possible nationwide. As hard equipment costs fell, customer acquisition and lead-generation “soft costs” became the biggest barrier, and Geostellar needed accurate building footprints for all U.S. counties to model rooftop solar at scale. Manually creating millions of vectors was infeasible, and the company required an automated, flexible image-analysis system able to handle varied LiDAR and imagery quality and reliably distinguish buildings from vegetation and other features.

Geostellar implemented Trimble eCognition object-based image analysis with custom rule sets to ingest LiDAR, satellite imagery and ancillary data, automatically extract building footprints, and feed its 3D solar-simulation engine. The automated workflow delivers huge time savings (up to ~90× versus manual processing) and enabled Geostellar to map about 80% of high-value solar markets at one-meter resolution. The platform now provides real-time rooftop solar assessments, financing and installer options, and powers initiatives like Connecticut’s GoSolarCT and the consumer SolarMojo app, helping reduce soft costs and scale solar adoption.

Geostellar

David Levine

Chief Executive Officer