OPC UA Pub/Sub

Enables direct, one-to-many communication on the local network without a central message broker. Publishers send UADP (UA Datagram Protocol) messages to a multicast group, and subscribers listen in.

• Zero broker infrastructure: Eliminates a central point of failure and reduces latency.
• Ethernet level efficiency: Uses standard Layer 2/Layer 3 multicast, ideal for high-speed machine-to-machine communication.
• Best-effort delivery: No handshake overhead, suitable for cyclic telemetry where occasional packet loss is acceptable.

Example: A single drive controller publishing actual torque values every 1ms to 20 subscriber nodes on a machine cell network.

Maps OPC UA data directly onto MQTT messages for cloud ingestion and wide-area distribution. Uses a central broker to manage subscriptions and ensure delivery.

• JSON encoding: Standard JSON payloads for direct consumption by cloud services like AWS IoT Core or Azure IoT Hub.
• UADP binary over MQTT: Compact binary encoding for bandwidth-constrained WAN links.
• Topic structure: Configurable MQTT topic namespaces derived from the OPC UA information model.

Example: A remote oil well controller publishing wellhead pressure data via MQTT to a central SCADA system over a satellite link.

Combines OPC UA Pub/Sub with IEEE 802.1 TSN to guarantee deterministic delivery over standard Ethernet. This is the foundation for real-time field-level control.

• 802.1Qbv Scheduled Traffic: Time slots are reserved for Pub/Sub frames, preventing collision with best-effort traffic.
• 802.1AS Precision Time Protocol: All publisher and subscriber clocks are synchronized to sub-microsecond accuracy.
• Bumpless redundancy: TSN frame replication and elimination (802.1CB) provides seamless failover.

Example: A motion controller publishing setpoint commands to 50 servo drives with guaranteed 31.25µs cycle time on a converged network carrying video and TCP traffic.

Defines a structured payload model that decouples the published data from the underlying transport protocol. A DataSet is a collection of DataSetFields representing individual variable values.

• DataSetMetaData: Describes the structure, names, and data types of the payload, enabling late-binding subscribers.
• Field-level encoding: Each field can use different data types (Int32, Float, String) within a single message.
• Versioned schema: DataSet versioning allows subscribers to detect and adapt to structural changes.

Example: A PublisherDataSet containing 'Temperature' (Float), 'Pressure' (Float), and 'ValveState' (Boolean) fields, with metadata published separately for dynamic discovery.

Implements message-level security independent of the transport, ensuring confidentiality and integrity across untrusted network segments.

• JSON Web Token (JWT): Authenticates publishers and subscribers to MQTT brokers.
• Sign & Encrypt: UADP messages can be signed for integrity and encrypted for confidentiality using asymmetric cryptography.
• SecurityGroup concept: A logical grouping of publishers and subscribers sharing a common security policy and key set.
• Push-based key rotation: Centralized Security Key Service (SKS) pushes new keys to all group members before the old key expires.

Example: A wind farm operator encrypting all turbine telemetry published over a shared carrier network, with keys rotated every 24 hours via SKS.

Subscribers consume data passively without establishing a session with the publisher. This fundamentally changes the scalability equation compared to client-server OPC UA.

• No session overhead: Publishers do not maintain state for each subscriber, enabling thousands of consumers.
• Late joining: Subscribers can join an active data stream at any time without publisher reconfiguration.
• Network decoupling: Publishers and subscribers do not need to know each other's IP addresses in brokered or multicast modes.

Example: Adding a new analytics dashboard that subscribes to an existing machine telemetry stream without any changes to the PLC publishing the data.

A decentralized, data-centric middleware protocol that shares the publish-subscribe paradigm with OPC UA Pub/Sub. While DDS provides a real-time data bus with rich Quality of Service policies, OPC UA Pub/Sub adds standardized information modeling.

• DDS excels in autonomous systems and defense applications.
• OPC UA bridges the gap between DDS's raw data and semantic context.
• Both can coexist in a Unified Namespace architecture.

A competing but complementary publish-subscribe specification for IIoT. Sparkplug defines a standard topic namespace and payload on top of MQTT, adding state management and birth/death certificates.

• Key Difference: MQTT Sparkplug is broker-centric; OPC UA Pub/Sub can be broker-less (UDP) or broker-based (AMQP/MQTT).
• Sparkplug is often used for cloud ingestion, while OPC UA Pub/Sub targets the deterministic edge.

A semantic data architecture that acts as the single source of truth for all industrial data. OPC UA Pub/Sub serves as the primary data injection mechanism into the UNS.

• Publishers push contextualized data directly into the namespace.
• Subscribers discover and consume data without point-to-point binding.
• The UNS eliminates the traditional automation pyramid's rigid hierarchy.

An event-driven, distributed control standard that naturally pairs with OPC UA Pub/Sub. IEC 61499 defines portable function blocks that communicate via events and data.

• Pub/Sub provides the ideal transport layer for function block event propagation.
• Together, they enable truly decentralized, hardware-agnostic control logic.
• This combination is the foundation of software-defined manufacturing.

Defined by IEEE 1588, PTP is the synchronization backbone that makes deterministic Pub/Sub possible. It distributes a master clock to all nodes on the network with nanosecond precision.

• Essential for coordinating distributed motion axes.
• Enables isochronous data sampling across multiple publishers.
• PTP is a prerequisite for TSN's time-aware scheduling.