- Support for time-series analysis so organizations can analyze device performance over time
- The storage of device data in a time series data store is optimized for quick response times to IoT queries that often include time among the criteria
- Tools for machine learning, supported hosted Jupyter notebooks. This allows you to connect IoT data directly to a notebook, build, train, and execute models right from the IoT Analytics console.
- Data preparation techniques support the setup and processing of data prior to analysis
- Automated scaling and pay-as you-go pricing
AWS stated that AWS IoT Analytics automates all the tedious steps required to analyze data from IoT devices. IoT Analytics transforms, enriches and filters IoT data before it is stored in a time-series storage for analysis. The service can be set up to only collect the data you need from your devices. It will then apply mathematical transforms to process the data and enrich it with device-specific metadata like device type and location. You can then analyze your data using the built-in SQL engine. Or, you can perform more advanced analytics and machine learning inference. The pricing details for the new service are available here.