Predicting carbon and cost savings of home efficiency projects

As the final project for a Terra.do Software Stacks in Climate Tech course, a group of climate-passionate Product Managers and Software Engineers workshopped this idea, created a proof-of-concept, and came up with several ideas for real-world applications. We're hopeful this will inspire others, and might even make its way into a smart thermostat someday!

Problem: Home heating is a huge contributor to climate change, and homeowners aren't doing enough.

High emissions from homes

40% of US emissions are from buildings, and 40% of that is from fossil fuels for heating

HVAC upgrades out of reach

Upstream changes can reduce the footprint of heating, like electrifying to use heat pumps powered by renewables, but these can be expensive or inaccessible

Untapped energy savings

Most homeowners haven't experienced the savings from home energy improvement projects

What if every homeowner could measure actual energy savings from home improvement projects, using data from thermostats they already have?

Our Theory of Change

Quantifying realized savings from home improvement projects (in terms of CO2e emissions and dollar cost savings) will create a positive consumer experience and lead average homeowners to invest in many more energy efficiency projects than they otherwise would.

Impactulator schematic

Impactulator schematic 2

Our Prototype

Sinceote we didn't have time to collect historical data using real thermostats, we used a basic home energy model to create a timeseries of fake data, which we used as the input for our ML model.