Background:

Our team was brought on to help utility companies unlock the power of big data and enable resilient energy grids. We recognized the need for a suite of machine learning models that could accurately forecast power loads one week ahead, as well as classify building electrification to mitigate weather driven outages. Our goal was to create an intelligent system that could handle situations with limited data and be used by utilities nationwide.

Problem:

The challenge we faced was developing a suite of ML models that could accurately predict power loads one week ahead. Additionally, we needed to develop a model that could classify building electrification in order to mitigate weather-driven outages. Due to the lack of available data, creating an accurate model posed a unique challenge.

Solution:

To meet these challenges, our team developed a comprehensive suite of ML models. First, we created an algorithm that leveraged historical data from utility companies across the country to accurately forecast power loads one week ahead. We then built a model that classified building electrification using satellite imagery and other environmental factors. Finally, we designed our system to be intelligent enough to handle situations with limited data.

Impact:

Accurately forecasts power loads one week ahead
Classifies building electrification to mitigate weather-driven outages
Intelligently handles situations with limited data

Team:

Rick Astley: Lead Engineer
- Developed algorithm leveraging historical data from utility companies across the country
- Built model classifying building electrification using satellite imagery and other environmental factors
: Data Scientist
- Designed system intelligent enough to handle situations with limited data

Creating Intelligent Energy Grids