Our approach is data-driven

We combine Big Data with machine learning and state of the art computer vision to predict who's most likely to adopt rooftop solar.

The use of state of the art computer vision

We identify all buildings with rooftop solar from publicly available satellite imagery. We determine the geo-location and estimate the number of panels.

Did you know that proximity has a huge influence on who goes solar? Showing the many others in a neighborhood who have gone solar, significantly increases the uptake of solar amongst new customers.

Calculate a roof’s solar potential

We use LiDAR (Light Detection and Ranging) to extract building related features and to estimate a roof’s PV suitability. By overlaying meteorological data we can accurately calculate a roof’s solar potential. Our approach allows us to automatically identify how many panels we can fit on a roof and optimises the panel placement accordingly.

Prediction model

A predictive sales platform

"One of the best predictors of people buying solar is knowing whether their neighbours did it first." (Kenneth Gillingham, a professor of Economics at Yale University.)

Combining Big Data with machine learning predicts what households are most likely to buy rooftop solar. Think of using demographics, income, property value, ownership and even social media to to determine a customer's preference and interest.

The better you understand your customer, the better you can tailor the customer experience and provide the right set of products and services.

Customer experience