Are you accurately forecasting energy usage?
Many people don't realize that retail energy companies rely heavily on their ability to forecast the energy that their customers will consume. Also, commonly referred to as the load, this forecast allows the providers to not only hedge energy prices, but it also brings with it the most important benefit of all - it enables them to provide consistent and predictable long-term pricing to their customers.
Recently, a major retail energy provider wanted to optimize this customer load prediction to enable even more accurate pricing than ever before. They wanted to be able to offer their customers the best energy prices possible, and Pariveda was more than happy to help them meet these needs.
Develop a cloud-based prediction platform
The hardworking team at Pariveda worked directly with the company to develop a fully customized cloud-based load forecasting solution. This solution used historical customer and environmental data in a linear regression model, allowing the company to better predict usage for an individual customer's meter rather than making predictions based on a series of customers that have been grouped together.
Processing time reduced from 8 hours to only 5 minutes
Thanks to Pariveda's solution, load forecasting processing time has been reduced from eight hours to just five minutes - generating a much more granular output of forecasting usage for each individual customer's meter at the same time. The company's customers now have a dramatically increased level of awareness of their future energy costs, and the retail energy company can provide more reliable prices from an accurate and granular load forecast, no exceptions.
Azure Data Lake Analytics and Azure Machine Learning Studio were used by Pariveda to power this essential linear regression algorithm, allowing the customer to maintain their competitive advantage in the marketplace for years to come.