Energy Usage Time-Series Forecast Application
Climate change is the significant issue of our time, so for the Capstone Project of the Le Wagon Bootcamp (Batch 903) my team decided to make a product focused around saving household energy. Our Energy Usage Prediction App, aims to help users manage their energy consumption, reduce carbon footprint and potentially save people money on their energy bills. In this post, I’ll examine the app, its features and illustrate how some of the web app functions.
Key Features of the Energy Usage Prediction App
Real World Data: The app is built on a dataset comprising smart meter readings from real houses and real people in the London area, providing accurate and reliable predictions.
Facebook's Prophet forecasting algorithm: The app uses Facebook's Prophet forecasting algorithm, which evaluates historical information to predict future energy usage. This algorithm has been specifically designed for such purposes and has been enhanced with local weather data and UK holiday information for increased accuracy.
Personalisation: By inputting information about your household and personal circumstances, the app provides a personalised energy recommendation tailored to your needs.
Easy access to individual reports: Your personalised energy report is just one click away, making it easy to access and understand your energy usage patterns.
How the App Works
The Energy Usage Prediction App works by asking users to input information about their household and personal circumstances, such as the type of house they live in, the number of bedrooms, and their annual salary. The app then uses this information to categorise users based on their socioeconomic status, as no two households consume energy in the same way.
Once the user is assigned to a category, the app combines the user's information with UK holiday and local weather data, along with historical energy consumption data from the same category. The app then uses either a machine learning model (Facebook's Prophet) or a deep learning model (Recurrent Neural Network) to generate energy usage predictions for the user.
The app displays a variety of metrics, including total monthly energy consumption, a comparison to the UK average, likely annual energy consumption, daily maximum usage, and the user's carbon footprint.
Potential Applications of the Energy Usage Prediction App
Individuals: The app is designed to help individuals manage and plan their personal energy consumption, allowing them to make informed decisions about their energy usage and ultimately reduce their carbon footprint.
Real Estate Sector: Integrating the app into real estate websites can provide potential buyers or renters with valuable information about the energy efficiency of a property, giving environmentally conscious consumers a competitive edge.
Government and Power Suppliers: On a national level, the app can be used to improve grid planning for increased efficiency, identify buildings with abnormal energy usage, and proactively address communities at risk of fuel poverty.
Conclusion
The Energy Usage Prediction App is an innovative solution to help individuals, real estate professionals, and governments better understand and manage energy consumption. By providing personalised energy recommendations based on real-world data, the app empowers users to make informed decisions about their energy usage, ultimately leading to a more sustainable and environmentally conscious future.
