SAS Energy Forecasting Cloud

From generation to distribution, get repeatable, traceable and defensible energy forecasts in the cloud. Scale up and down depending on the requirements of your business.

Streamline and automate your energy forecasting process

SAS Energy Forecasting Cloud enables you to automate data ingestion and statistical and AI/ML model building so you can select the best one without the need for data science expertise. Whether you are a dispatcher or a trader, you can make the timeliest operations decisions and market offers, like never before.

Get accurate, robust energy forecasts at your fingertips

SAS Energy Forecasting Cloud systematically considers hundreds of combinations of factors to develop the best model to use for your short- and very-short-term forecasting needs. As a result, robust forecasts are always available in the control and trading rooms. In fact, our customers have consistently seen mean absolute percentage errors (MAPEs) of 2% or less*. *using quality regional-level data.

Create load and renewable generation forecasts for an expanding energy grid

Get hourly and sub-hourly forecasts for load and renewable generation, enabling your organization and your customers to have accurate and reliable net-load forecasts that help optimize the use of energy resources.

Ensure compliance with defensible and scalable forecasts

Be prepared for any situation with defensible and scalable forecasts. With SAS Energy Forecasting Cloud, you can quickly and accurately generate and defend your forecasts with respect to regulatory compliance. Our service forecasts thousands of time series and saves all data, models and results until they are archived.

Key features

Automation, scalability, statistical sophistication and transparency for operating more efficiently and effectively at all levels of decision making.

As a service

Delivers the quality energy forecasts you’ve come to expect from SAS Energy Forecasting without having to maintain software in your facilities. Lets you scale up and down as your business demands.

Load forecasting

Automates and streamlines data ingestion for very-short-term and short-term forecasting, including these models: GLM, ARIMA, NN, UCM, ESM.

Renewables forecasting

Provides an automated, scalable solution for traditional load forecasting and extends the capabilities to effectively model renewable generation resources, such as solar and wind, with advanced machine learning and deep learning algorithms – ultimately producing a net load forecast that supports the needs of expanding diverse energy grids.

Advanced forecasting algorithms

Enables hourly and sub-hourly forecasting based on trusted data and advanced forecasting algorithms. Ensures reliable data with data quality capabilities.

High-performance load forecasting

Enables utilities to operate more efficiently by maximizing value from existing planning resources and improving forecast performance.

Single administration & reporting interface

Provides a visual interface for viewing forecasting results from the forecast workbench. Autocharting capabilities mean no coding is required.

Flexibility & scalability

Automated champion model selection and forecasting are available as a service. The completely redesigned architecture is compact, cloud native and fast.

Improved load forecasting performance & operations planning capabilities

Improves forecasting across all locations and levels of aggregation with repeatable, scalable and traceable results.

Better trading & contract purchase decisions

Lets modelers incorporate quantifiable variability and confidence limits when making operational and financial decisions through statistical and visual indications of the likely range of forecasted outcomes.

Maximum ROI from smart meters & advanced metering infrastructure using all data

Makes better predictions about energy demand possible with accurate predictive models based on data from more sources, including smart meters and IoT-connected devices.

Ability to do more with existing planning & forecasting resources

Eliminates the need to train forecasters on multiple software tools via a common forecasting methodology and data integration processes across forecasting horizons.