A large manufacturing company wants to implement a new IoT solution for factory automation. Thousands of sensors will generate high-volume, high-velocity time-series data (e.g., temperature, pressure, vibration) from machines across multiple factories. This data needs to be ingested, stored for long-term analysis, and used for real-time anomaly detection and predictive maintenance. The solution must be scalable, cost-effective, and provide near real-time insights. Which architecture should be designed?
- AAWS IoT Core for device connectivity, Amazon Kinesis Data Streams for ingestion, Amazon DynamoDB for time-series storage, and Amazon Athena for ad-hoc analysis.
- BAWS IoT Core for device connectivity, Amazon Kinesis Data Firehose for ingestion, Amazon Timestream for time-series data storage, and AWS Lambda with Amazon SageMaker for real-time anomaly detection.
- CAWS IoT Core for device connectivity, Amazon Kinesis Data Streams for ingestion, Amazon S3 for long-term storage, and AWS Lambda for processing.
- DAWS IoT Core for device connectivity, AWS IoT Analytics for data processing and storage, and Amazon QuickSight for visualization.
Show answer & explanationAnswer & explanation
Correct answer: B. AWS IoT Core for device connectivity, Amazon Kinesis Data Firehose for ingestion, Amazon Timestream for time-series data storage, and AWS Lambda with Amazon SageMaker for real-time anomaly detection.
This architecture leverages AWS IoT Core for device management, Kinesis Data Firehose for efficient ingestion into Timestream, which is purpose-built for time-series data storage and analysis. Lambda with SageMaker provides the necessary real-time anomaly detection and predictive maintenance capabilities, meeting all requirements for scalability, cost-effectiveness, and near real-time insights.
Why the other options are wrong
- A. DynamoDB can store time-series data but is not purpose-built for it like Timestream, potentially leading to higher costs and less efficient queries for this specific data type. Athena is for ad-hoc analysis, not real-time anomaly detection.
- C. While a valid pattern, S3 is not optimized for time-series queries, and Lambda alone might not be sufficient for complex real-time anomaly detection and predictive maintenance without additional services.
- D. AWS IoT Analytics is good for IoT data, but Amazon Timestream is specifically optimized for time-series data storage and query performance, which is a key requirement. QuickSight is for visualization, not real-time anomaly detection.
IoT Time-Series Data Platform
An AWS architecture for ingesting, storing, and analyzing high-volume, high-velocity time-series data from IoT devices, enabling real-time insights and predictive maintenance.
- Uses AWS IoT Core for device connectivity and message routing.
- Leverages Kinesis for data ingestion.
- Employs Amazon Timestream for optimized time-series storage.
- Integrates Lambda/SageMaker for real-time analytics and ML.
Memory trick: IoT Core Streams to Timestream for ML Insights.