GridGenie

Inspiration

Renewable-powered microgrids bring together solar generation, wind energy, battery storage, facility demand, and grid connections. Understanding how these components interact, especially when conditions change, can be challenging. We wanted to make microgrid operations easier to explore and energy-management decisions easier to evaluate.

What it does

GridGenie provides an interactive 3D digital twin of a microgrid, allowing users to explore its assets and understand how different operating conditions affect the system. Users can investigate operational scenarios, compare strategies, examine energy forecasts, and explore optimization-driven decisions.

The platform combines a visual microgrid environment with machine-learning-based forecasting and mathematical optimization, bringing system visualization and analytical decision support into one interface.

How we built it

GridGenie uses React and React Three Fiber for its interactive frontend and 3D environment, with a Python and FastAPI backend supporting its analytical functionality. Scikit-learn Ridge regression is used for forecasting, while a linear programming optimization pipeline uses SciPy's HiGHS solver to evaluate operating decisions.

Challenges we faced

One of the central challenges was bringing 3D visualization, backend analytics, forecasting, and optimization together into a coherent application. The interface needed to make energy-system behavior understandable while keeping the analytical outputs accessible and useful.

What we learned

Building GridGenie strengthened our understanding of integrating interactive scientific visualization with machine learning and mathematical optimization. It also reinforced the importance of presenting analytical results in a form that people can explore and interpret.

What's next

Future improvements could include more detailed energy-system modeling, additional renewable-energy scenarios, stronger forecast validation, and expanded operational constraints to support more realistic microgrid planning and decision-making.

Built With

• fastapi
• highs
• python
• react
• scikit-learn
• scipy
• three.js
• vite