Our project, SolarSense, is a machine learning model that predicts various solar indicators to optimize solar production by finding the most efficient places to utilize for solar energy. We trained a model to predict the global horizontal irradiance (GHI), the diffuse horizontal irradiance (DHI), and the direct normal irradiance (DNI) based on various environmental factors such as time, humidity, temperature, cloud covering, and the solar zenith angle. We visualized predictions for various locations on our custom interactive map and globe found here: https://www.google.com/maps/d/u/0/edit?mid=1Hlr7-YINhevdlAi4JVibfdtPNkffovE&usp=sharing

Direct Normal Irradiance (DNI) measures the amount of solar radiation that hits a surface that is always facing directly towards the sun. This type of radiation does not get scattered by the atmosphere and is therefore the most intense type of solar radiation. Diffuse Horizontal Irradiance (DHI), on the other hand, measures the amount of solar radiation that arrives at a surface after being scattered by particles and molecules in the atmosphere. This type of radiation comes from all directions and is not as intense as direct radiation. Global Horizontal Irradiance (GHI) measures the total amount of solar radiation that reaches a surface horizontal to the ground. This value is important for photovoltaic installations, as it includes both direct and scattered radiation, and is therefore a key factor in determining the energy output of a solar panel.

SolarSense greatly simplifies the process of determining the optimal location for solar panel placement. By automatically predicting the optimal heat environments for solar panels, the SolarSense model can help maximize renewable energy production and minimize waste. This will lead to increased efficiency and reduced costs for solar energy installations, improving sustainability on multiple fronts and fostering positive environmental change.

SolarSense enables anyone to easily predict the potential effectiveness of solar cells in a given area at no cost. With the large amounts of available weather data and virtually no operational cost, SolarSense is a low-cost solution to an impactful sustainability issue. In addition to this, SolarSense can be easily maintained and scaled with new data by just training the model again, allowing it to continue to facilitate efficient solar planning for the future.