GridSense

Inspiration

As college roommates, we wanted to make our apartment more sustainable without ignoring the reality of our electricity bill; these two factors go hand in hand. We built GridSense to help everyday households shift energy use into cleaner hours, reduce avoidable emissions, and make smarter decisions about when to run high energy-consumption appliances.

What it does

For a specified location, GridSense shows real-time grid carbon intensity, forecasts cleaner and dirtier energy windows over the next 24 hours, predicts periods of grid stress, and recommends better times to use appliances like EV chargers, washers, dryers, and dishwashers. The goal is to help users lower their carbon footprint by aligning energy use with cleaner grid conditions.

How we built it

We built the dashboard utilizing a FastAPI backend and a React + Vite frontend. We used WattTime for grid emissions data, Open-Meteo for weather signals, Google Maps for city geolocation and map context, and Azure OpenAI to generate concise, actionable appliance nudges based on forecasted clean-energy windows.

Challenges we ran into

One challenge was combining multiple live data sources into a single reliable user experience. We also had to handle provider failures, format real-time and forecast data clearly, and design visualizations that made carbon intensity and grid stress easy to understand at a glance. This took a lot of iterating and redefining our values on what we wanted for our end product.

Accomplishments that we're proud of

We are proud that our project turns complex sustainability data into practical actions people can actually follow, presented in a simple and easy to understand manner. We also built a full working flow from emissions data to forecast visualization to appliance recommendations, making implementing sustainability feel convenient and useful.

What we learned

We learned how difficult it is to transform raw climate and energy data into a product that is both technically credible and easy to use. We also learned a lot about API integration, fallback handling, frontend data visualization, and how AI can support sustainability decisions when grounded in real environmental signals.

What's next for GridSense

Next, we want to personalize recommendations based on household routines, estimate actual cost and carbon savings from shifted usage, expand location coverage, and make the platform more predictive so users can plan sustainable energy use further in advance. Basically, make the dashboard more customizable and tailored for each individual household.

Built With

• azureopenai-api
• css
• googlemaps-api
• html
• javascript
• node.js
• python
• watttime