Esri, Microsoft, and Synthetaic Partnership

The data science team at Chesapeake Conservancy developed an AI model that maps ground-mounted solar arrays in satellite imagery. They used this model and the Microsoft Azure and the Planetary Computer to create annual maps of all solar arrays in the Chesapeake Bay area from 2017 to 2022.

In the last several years, Synthetaic has evolved the traditional AI, machine learning, and deep learning workflows to greatly streamline the process of labelling, training and detecting objects. Their new approach, which runs entirely on Microsoft Azure with datasets contained in the Planetary Computer, allows the Chesapeake Conservancy to detect ground-based solar panels on higher resolution imagery (NAIP) efficiently, reducing the amount of time spent manually labeling training data.

Synthetaic can also use their system to extract additional features, such as poultry farms, that can provider greater insights into the activity in and around the Chesapeake Bay.

When this workflow is complete, further analysis can be done by Esri’s ArcGIS Pro tools running within Microsoft Azure connected to the Planetary Computer. The analysis creates additional datasets using Esri’s pre-trained AI models along with other solutions, bringing together additional contextual information from existing layers hosted by CC and other stakeholders in the region.

The data was pulled together to provide a dynamic view of the affected area for situational awareness.