Episode Details

When a machine stops, a planner needs more than a red line on a dashboard. They need to know what stopped, why it stopped, which order sits on that line, and whether the production plan can still hold. Moving raw signals from a factory floor into cloud analytics requires a reliable architecture that spans field sensors, industrial gateways, message brokers, and cloud ingestion layers.

CAPTURING THE SIGNAL AT THE EDGEA sensor signal is not yet manufacturing information. Picture a packaging line where a conveyor motor temperature rises beyond normal operating limits. At that moment, the programmable logic controller (PLC) holds a raw tag value—such as a temperature reading, a timestamp, and a quality state.

However, older plants often run multiple generations of equipment using proprietary protocols or OPC UA, meaning PLCs rarely speak cloud MQTT directly. An industrial gateway deployed on the operational technology (OT) network bridges this gap by:

• Normalizing Data: Translating disparate formats, scales, and engineering tags into a common message shape.
  
• Managing Collection Frequency: Aggregating high-frequency scans into meaningful intervals while forwarding state changes immediately.
  
• Buffering Locally: Storing data securely during network outages and replaying it safely once connectivity returns.
  

MOVING MESSAGES WITH MQTTOnce the gateway shapes a message, it publishes it using MQTT (Message Queuing Telemetry Transport), a lightweight publish-subscribe protocol well-suited for constrained devices and unreliable links.

• Publish-Subscribe Decoupling: Publishers send messages to a known topic without needing to know which downstream systems or applications consume them.
  
• Topic Structure: Topics act as transport addresses for categories of messages, but they should not be overloaded to serve as a full factory asset model.
  
• Payload Contracts: The actual message body must explicitly carry stable source asset identifiers, measured values, units, timestamps, and quality flags.
  

THE ROLE OF AZURE IOT HUBWhile local MQTT brokers handle internal plant traffic, Azure IoT Hub acts as the managed cloud connection layer for enterprise fleets.

• Device Identity & Authentication: Every gateway requires a secure, registered device identity in IoT Hub using either Shared Access Signatures (SAS) or X.509 certificates.
  
• Transport Security: Connections rely on TLS (Transport Layer Security) to protect data in transit across network boundaries and firewalls.
  
• Built-in Endpoints: IoT Hub accepts device-to-cloud messages, stamps them with trusted connection metadata, and retains them temporarily before routing them onward.
  

ROUTING, STREAMING, AND LONG-TERM STORAGEIncoming telemetry serves multiple consumers across the organization, requiring message routing to split traffic into distinct paths:

1. Raw History Storage: Routing messages to Azure Blob or Data Lake Storage preserves an unaltered audit trail for engineering reviews and reprocessing.
   
2. Stream Processing: Endpoints like Event Hubs feed near-real-time streaming consumers and alert rules to notify maintenance teams of abnormal conditions.
   
3. Analytical Models: Integrating data streams into platforms like Microsoft Fabric helps combine telemetry with enterprise planning and execution context.
   

Become a supporter of this podcast: https://www.spreaker.com/podcast/m365-fm-a-microsoft-mvp-podcast-by-mirko-peters--6704921/support.