When considering the installation of solar panels, a common question arises: “How many panels are needed?” Similarly, for heat pumps, the question is “What size is required?” Typically, answering these questions necessitates a visit from a contractor to the property in question.

However, over the past decade, solar installers have significantly streamlined this time-consuming process. Companies like Tesla have even eliminated the need for on-site visits entirely. The key to this efficiency lies in the utilization of software, which, when combined with various data sources, enables installers to accurately model the number of panels a house requires.

Similar to other solar installers, Palmetto has developed its own internal tool for assessing potential projects. This tool is built on software acquired when Palmetto purchased Mapdwell, a solar mapping startup.

As the market for residential solar has experienced slowed growth, Palmetto has shifted its focus to other electrification projects, including backup batteries and heat pumps. Developing software for these products, which are located inside homes and thus not visible through satellite imagery, presents a more complex challenge.

To address this, Palmetto has initiated an effort to gather data for creating a virtual representation of every residential building in the United States. According to Michael Bratsafolis, Palmetto’s president of energy intelligence, the company is “essentially simulating digital twins of the entire US residential building stock.”

The company’s tool relies on both public and private data sources. In instances where exact information is unavailable, the tool’s AI utilizes contextual clues to infer the characteristics of a building. For example, a house constructed in the 1950s likely features 2×4 lumber for exterior walls, which limits the amount of insulation that can be installed.

“This technology can break down a home into more than 60 different characteristics and attributes,” Bratsafolis explained.

To validate the models, Palmetto can utilize the data it has gathered from homes that have purchased or leased solar panels from the company.

“This allows us to anonymize and use real information and data about a home to compare how the model is performing and help train the model further,” he said.

Palmetto has decided to make the tool accessible to outside developers, a move inspired by Bratsafolis’ experience at Twilio.

“I came from the developer motion, the micro-service approach to providing API building blocks to empower partners and developers,” he said.

Developers are allocated 500 free calls per month, and subsequent calls are charged at five cents each. For larger customers, Bratsafolis indicated that the company is willing to negotiate special pricing.

Bratsafolis hopes that developers will create tools that accelerate the electrification of US building stock, and Palmetto will gain early insights into market trends: “You don’t necessarily know what the use cases could be. And with an API product, you gain incredible insight into demand and the market.”