Energy Impact
CLIENT
A Content Management System (CMS)
SERVICES
Data Collection and Integration, Advanced Data Analytics and Visualization, AI-Powered Energy Forecasting.
DATE
22 July 2021
Introduction
In a recent strategic meeting in London with Conrad Phillips, Director of Energy Impact, we engaged in in-depth discussions and talent acquisition for upcoming innovative projects. The details of these initiatives are confidential at this stage; however, the primary focus is the transformative approach to forecasting energy usage within the corporate sector.
Embarking on this journey, our team faced the formidable task of accumulating, storing, and analyzing electric usage data on a half-hourly basis for various entities including companies, individual sites, and meters. One significant hurdle involved the seamless procurement of this data from smart meters and its subsequent integration into our database, all while maintaining a vigilant eye for any discrepancies or data anomalies.
Triumph came in the form of an innovative dashboard, a testament to our team's hard work and perseverance. This tool efficiently displays years of data, allowing for swift, comprehensive comparisons of energy usage across diverse time frames. The utility of this dashboard is further amplified by its ability to present data in a multitude of visual formats, including line, area, bar, and radar charts, thereby broadening our analytical capabilities.
Our journey didn't halt at data collection and visualization; it evolved with the integration of Artificial Intelligence. Our ambitious goal: to meticulously analyze three years of historical data to predict future electricity consumption patterns. The complexity of this task was compounded when dealing with multifaceted businesses like football clubs, where a plethora of variables—from pandemic effects to diverse events—play a substantial role. Initial AI modeling led us to a commendable 74% prediction accuracy. However, our relentless pursuit of excellence didn't allow us to stop there. Delving deeper into hourly consumption patterns and identifying specific trends and anomalies, we recalibrated our focus and efforts toward predicting hourly usage— a move that has shown immense promise.
As we transition into the forthcoming phase of our project, our sights are set on the integration of more nuanced variables, for instance, understanding the ripple effect of LED lights on consumption patterns. Our ultimate aspiration is to meticulously hone our system's predictive prowess by embracing these additional dimensions of data. We owe a debt of gratitude to Energy Impact Limited for not only being an incredible partner on this journey but also for their generosity in allowing us to narrate our shared story. Their team's spirit and dedication are unparalleled. Stay connected for more insights and updates as we navigate this captivating path of innovation and discovery!
