SAS® Analytics for IoT Features
Streamlined, extensible ETL
- Automatically transforms and loads key data fields into the sensor-based data model.
- Lets you rapidly load IoT data, whether you have three fields (sensor ID, value and date time) or hundreds.
- Includes sensor attributes, device attributes, hierarchies, measures and events.
- Enables integration of additional field and production quality data with your sensor data, using comprehensive ETL capabilities.
Flexible, sensor-focused data model
- Provides a standardized, extensible sensor-based data model.
- Stores complex IoT data, hierarchies and other relationships right out of the box.
- Organizes large volumes of diverse IoT data for efficient analysis.
- Provides a single version of the data for diverse users across the organization.
Unified, intuitive business-focused data selection user interface
- Lets nontechnical users quickly select data for analysis without knowledge of the underlying technology and data structure.
- Enables users to access available variables and attributes in their own business terminology.
- Uses smart filters, predefined date windows and other shortcuts to increase efficiency and reduce errors.
- Supports individual user needs by letting them select data for any combination of devices, sensors, measures and events.
- Allows you to save, copy, reuse and share data selections across the organization.
- Enables users to easily prepare and transform data for analysis in SAS or third-party tools.
- Transposes data from an efficient storage format to an analytics-ready format.
- Interpolates missing values in the data.
- Applies a fixed periodicity to reduce data size or commonize across sensors.
- Lets users open the data in SAS Visual Analytics, SAS Visual Data Mining and Machine Learning, and SAS Studio, as well as third-party and open source applications.
Advanced analytics and machine learning
- Combines data exploration, feature engineering and modern statistical, data mining and machine learning techniques in a single, scalable in-memory processing environment.
- Lets users analyze data without writing code, using a drag-and-drop interactive interface.
- Uses best practice templates (basic, intermediate or advanced) to get you started quickly with machine learning tasks.
- Applies diverse machine learning algorithms – including decision trees, random forests, gradient boosting, neural networks, support vector machines and factorization machines.
- Compares results of multiple machine learning algorithms with standardized tests to automatically identify champion models.
Streaming model execution
- Analyzes and filters streaming data (data in motion) in real time.
- Lets you create, deploy and manage advanced analytics models running on streaming data.
- Scores data in real time and applies learning models that combine scoring and training.
- Reduces downstream processing by cleansing, standardizing and filtering livestream data before it’s stored.
- Allows external systems to access data in a way that optimizes IoT investments across the enterprise.
- Lets you integrate SAS or third-party solutions into your IoT ecosystem.
- Automatically populates external dashboards and reports with the latest data or lists of data selections.