What is Azure IoT Operations?

2025-08-21

Azure IoT Operations is a unified data plane for the edge, offering modular and scalable data services on Azure Arc-enabled Kubernetes clusters like AKS Edge Essentials. This article explores its features, benefits, and use cases.

Azure IoT Operations:

Is built from the ground up using Kubernetes-native applications.
Is part of the Microsoft adaptive cloud approach that unifies siloed teams, distributed sites, and disparate systems into a single operations, security, application, and data model.
Includes an industrial grade, edge-native MQTT broker that powers event-driven architectures.
Is highly extensible, scalable, resilient, and secure.
Can operate offline for a maximum of 72 hours. Degradation might occur during this period. However, Azure IoT Operations resumes full functionality when it reconnects.
Lets you manage edge services and resources from the cloud by using Azure Arc.
Enables secure management of devices in layered networks using open, industry-recognized software, and Kubernetes-based configuration.
Can integrate customer workloads into the platform to create a unified solution.
Natively integrates with Azure Event Hubs, Azure Event Grid's MQTT broker, and Microsoft Fabric in the cloud.

Use Azure IoT Operations to:

Improve business efficiency and decision-making by using AI in the cloud to analyze asset and equipment data from the edge. Azure IoT Operations processes and normalizes the data at the edge before it's sent to the cloud.
Transform manufacturing environments by removing barriers between OT and IT systems. Azure IoT Operations supports open standards such as MQTT and OPC UA, and frameworks such as Kubernetes that enable it to foster interoperability and run processes like predictive maintenance, energy optimization, and digital inspection.
Modernize on-premises and edge infrastructure to handle digital operations. Azure IoT Operations offers a suite of services that enable you to connect, manage, and receive data from your assets.
Secure your end-to-end operations by using Azure security capabilities. Azure IoT Operations includes built-in security features like secrets management, certificate management, and secure settings.

Example use cases

Use Azure IoT Operations to address use cases such as:

Anomaly detection

To identify anomalies in the data generated by an industrial asset, an operator can use the operations experience web UI to:

Connect an OPC UA asset to the Azure IoT Operations MQTT broker at the edge.
Define a data flow that processes and normalizes the data before identifying any anomalies.
Send the processed data to Microsoft Fabric in the cloud.

Use Microsoft Fabric to build real-time dashboards with visualizations that show the status of the asset and alerts for detected anomalies. You can make these dashboards available on the shop floor where operators can use them to take immediate action and mitigate potential issues. Using predictive analytics and data on the edge helps anticipate failures before they occur and reduces downtime and maintenance costs.

Operational equipment effectiveness

With Azure IoT Operations, you can use data collected from assets and equipment to improve your operational equipment effectiveness. Azure IoT Operations captures real-time data at the edge and processes it, enabling monitoring of key performance indicators such as availability, performance, and quality. Use Azure IoT Operations to normalize and analyze the data to identify patterns and areas for improvement.

Architecture overview

Azure IoT Operations architecture has two core elements:

Azure IoT Operations: A set of data services that run on Azure Arc-enabled edge Kubernetes clusters. These services include:
- MQTT broker to power event-driven architectures as an edge-native MQTT broker.
- Akri connectors, like the connector for OPC UA, to simplify communication with servers and leaf devices.
- Data flows to transform and contextualize data, letting you route messages to various locations, including cloud endpoints.
Operations experience: A web UI that lets operational technology (OT) users manage assets and data flows in an Azure IoT Operations deployment.

Manage devices and assets

Azure IoT Operations connects to various industrial devices and assets. Use the operations experience or the Azure CLI to manage the devices and assets you want to connect to.

Azure IoT Operations uses the Azure Device Registry to store information about local assets in the cloud. The service lets you manage assets on the edge from the Azure portal or the Azure CLI. The Azure Device Registry uses namespaces (preview) to organize assets and devices. Each Azure IoT Operations instance uses a single namespace for its assets and devices. Multiple instances can share a single namespace.

The Azure Device Registry includes a schema registry for assets. Data flows use these schemas to deserialize and serialize messages.

Discover devices and assets

The Akri services can discover devices and assets automatically, reducing the configuration overhead for OT users. OT users can use the operations experience web UI to view and manage discovered devices and assets.

The MQTT broker runs on the edge and lets you publish and subscribe to MQTT topics. Use the MQTT broker to build event-driven architectures that connect devices and assets to the cloud.

Examples of how components in Azure IoT Operations use the MQTT broker include:

The connector for OPC UA publishes data from OPC UA servers and leaf devices to MQTT topics.
Data flows subscribe to MQTT topics to retrieve messages, process them, and send them to cloud endpoints.

Process data

Data flows provide data transformation and contextualization capabilities within Azure IoT Operations. OT users can use the operations experience web UI to create and manage data flows.

Connect to the cloud

To connect to the cloud from Azure IoT Operations, you can use data flow destination endpoints like:

Visualize and analyze sensor data

To visualize and analyze sensor data and messages from your devices and assets, use cloud services like:

Next step

Try the Quickstart: Get started with an end-to-end sample.