Stream Analytics
Definition
Azure real-time analytics service for processing streaming data. Like having a smart analyst that can spot patterns and trends in live data streams.
Use Cases
- Rolls-Royce: Aircraft engine health monitoring using continuous telemetry to detect anomalies and predict maintenance needs — Streams of sensor telemetry are ingested into Azure and processed in near real time with Azure Stream Analytics to compute rolling aggregates, detect threshold breaches, and route alerts and enriched events to downstream stores and dashboards (Faster detection of abnormal operating conditions and improved maintenance planning, helping reduce unplanned downtime and operational risk)
- Komatsu: Monitoring heavy equipment telemetry to improve fleet operations and maintenance — Equipment sensor data is streamed to Azure and analyzed with Azure Stream Analytics to generate near-real-time operational metrics and trigger notifications when patterns indicate potential issues (More timely operational insights and improved ability to respond to equipment conditions, supporting better uptime and service outcomes)
Provider Equivalents
- AWS: Amazon Kinesis Data Analytics
- Azure: Azure Stream Analytics
- GCP: Google Cloud Dataflow
- OCI: OCI Streaming + OCI Data Flow
Frequently Asked Questions
- What's the difference between Azure Stream Analytics and Azure Data Factory?
- Azure Stream Analytics is for real-time processing of events as they arrive (seconds to minutes), using streaming queries and time windows. Azure Data Factory is for orchestrating data movement and batch ETL/ELT workflows (minutes to hours), scheduling pipelines and coordinating activities across services.
- When should I use Azure Stream Analytics?
- Use it when you need near-real-time insights or actions from continuous event streams—like IoT sensor monitoring, fraud detection, clickstream analytics, operational dashboards, or alerting. It’s a good fit if you want a fully managed service with SQL-like queries and built-in windowing, and you’re processing data from sources like Event Hubs, IoT Hub, or Kafka-compatible inputs.
- How much does Azure Stream Analytics cost?
- Pricing is primarily based on the number of Streaming Units (SUs) you allocate (which represent compute/memory for the job) and how long the job runs. Costs can also be influenced by features and data movement (for example, reading from Event Hubs/IoT Hub and writing to sinks like Blob Storage, Data Lake, Synapse, or Power BI may incur separate service charges). To estimate accurately, choose an SU level, expected runtime, and account for input/output service costs.
Category: data
Difficulty: advanced
Related Terms
See Also