AWS IoT Analytics End of Support - Migration Guide for IoT Data Analytics Pipelines
AWS IoT Analytics reached end of support on December 15, 2025. Explains the official migration targets (Kinesis Data Streams / Data Firehose, S3 / Timestream, Glue + Athena) and the 4-component architecture used while the service was offered.
End of Support and Migration Targets
AWS IoT Analytics reached end of support on December 15, 2025. Before that, AWS had stopped accepting new customers on July 25, 2024. You can no longer access the console or the service's resources. The official AWS migration guide presents the following alternative services for each function. Data collection is Amazon Kinesis Data Streams or Amazon Data Firehose, and transformation and processing is Amazon Managed Service for Apache Flink or Data Firehose + Lambda. Storage is Amazon S3 or Amazon Timestream, and analytics is AWS Glue and Amazon Athena. This page is kept as a historical record of the specifications at the time the service was offered + a migration guide.
IoT Data Analytics Challenges and the Role of IoT Analytics
IoT devices continuously send large volumes of telemetry data (temperature, humidity, vibration, location information, operating status, and more). Analyzing this data requires preprocessing such as removing noise, filling in missing values, converting units, and filtering outliers. An architecture in which IoT Core receives the data, Lambda preprocesses it, it is stored in DynamoDB or S3, and Athena queries it is possible, but you need to build and manage the integration between each component yourself. AWS IoT Analytics provided the pipeline from IoT data collection to analysis as a managed service. It consisted of 4 components: channels (data ingestion), pipelines (preprocessing), data stores (storage), and datasets (query results), and it could automatically process, accumulate, and analyze data from IoT Core.
The 4-Component Architecture
A channel is the entry point that receives data from IoT Core rule actions or the BatchPutMessage API. Received raw data is stored as is and can also be used for reprocessing. A pipeline retrieves data from a channel and applies a series of activities (processing steps). Built-in activities include adding and removing attributes, filtering (excluding data that does not match conditions), mathematical transformations (unit conversion), and adding metadata from the device registry. You can also run custom preprocessing logic with a Lambda activity. A data store is the storage that accumulates the data processed by a pipeline. It uses an S3 bucket as its backend and also supports storage in Parquet format. If you set a retention period, old data is deleted automatically. A dataset stores the results of a SQL query and can be updated regularly through scheduled runs (hourly, daily, and so on).
Analytics and Visualization
SQL queries in datasets run analyses such as aggregation, filtering, and joins against the data in a data store. For example, you can define queries such as "the average and maximum temperature of each device over the past 24 hours" or "a list of devices where outliers (threshold exceeded) occurred," and update the results regularly with scheduled runs. You can connect dataset results directly to Amazon Quick Sight (formerly Amazon QuickSight) to build dashboards. They visualize device operating status, sensor value trends, anomaly detection alerts, and more in real time. With the Jupyter Notebook integration, you can access data store data directly from a SageMaker notebook instance and build and validate ML models. It can be used for ML use cases such as predictive maintenance (equipment failure prediction), anomaly detection, and demand forecasting. There is also a feature that runs containerized analysis code as a dataset action, which lets you build periodic ML inference pipelines.
Pricing and Choosing Between IoT Analytics and Timestream
With the end of support, the following pricing structure no longer exists (a record from the time the service was offered). IoT Analytics pricing was $0.20 per 1 million messages for message processing (pipelines), $0.03 per GB per month for data storage, and $5.00 per TB of data analyzed for queries.
Data Quality and the Importance of Preprocessing
Data gathered from IoT devices often contains missing values, outliers, and inconsistent formats, and cannot be used for analysis as is. IoT Analytics let you configure a pipeline that performs preprocessing such as ingesting raw data and then excluding unneeded data, filling in missing values, and adding external information to give the data meaning. This let you prepare well-organized datasets suited for analysis. Because the accuracy of analysis depends heavily on the quality of the input data, this preprocessing stage determines the results. Shaping raw sensor data into a reliable foundation for analysis is the prerequisite for drawing useful insights from IoT data.
Summary - Migration Guidelines
AWS IoT Analytics was a service that achieved collection, preprocessing, and analysis of IoT device data with a managed pipeline, but it reached end of support on December 15, 2025. When you build an equivalent analytics platform, the basic form, as the official migration guide shows, is an architecture that combines the following services. Collection is handled by Kinesis Data Streams / Data Firehose, preprocessing by Managed Service for Apache Flink or Data Firehose + Lambda, storage by S3 / Timestream, and analytics by Glue + Athena. The design principle of this article, ensuring data quality through preprocessing, remains valid as is in the post-migration architecture.
References (Official AWS Resources)
The primary sources for this page are the official AWS website and documentation. Check the official pages below for the latest specifications and pricing.
- AWS IoT Analytics official page
- AWS IoT Core official page
- Amazon Quick official page
- AWS Documentation (official)
If this page and the official documentation disagree, treat the official documentation as authoritative.