November 12, 2020
What’s best for IIoT: An integration hub or an MQTT broker?
Complexity of IIoT infrastructure, needs should determine choice
In the white paper “Data is the key in IT/OT convergence” we discuss how the Industrial Internet of Things (IIoT) and Industry 4.0 are changing the way organizations are doing business by bridging the gap between operational technology (OT) and information technology (IT). One avenue is through data integration, and below, we provide a quick overview of what you should look for when selecting the right tool for integrating IT and OT data.
MQTT is an important building block of machine-to-machine communications. A main reason for this is because it is simple to use and is designed for use with constrained, low-bandwidth devices. MQTT enables IIoT devices to send sensor data, machine status and alerts, while requiring minimal code writing. It also allows equipment to receive commands that control behavior.
An MQTT broker is the middleware that manages message routing within and across factories and edge computing environments. But if you have a complex environment, such as a manufacturing floor, an energy production facility, etc., you will need more than a simple MQTT broker to implement manufacturing feedback loops, fully utilize IT/OT convergence benefits, and leverage machine learning, big data analytics, etc.
Taking integration to the next level
In basic terms, OT produces and uses data to control systems, and IT manages data. And in most cases, the IT and OT systems use different languages, protocols, applications, etc., making integration difficult. This is precisely where an integration hub becomes valuable.
A robust integration hub should be able to work with multiple protocols and file formats. On the OT side this includes MQTT, OPC, TCP, JSON, XML, Images, PDFs, etc. And for IT interests, this would include Message Queues (JMS and AMQP), REST, JSON over HTTP, PTC/ThingWorx, CSV files, SQL ODBC, SQL JDBC, etc.
In addition, you should also seek out an integration hub that has robust store and forward data capabilities to guarantee message delivery when subscribers are unavailable. It should replicate its data throughout a mesh network of edge devices to ensure extreme availability, and be able to receive data from all major IIoT data protocols and formats. It should deliver IIoT data to IT using the preferred data protocols and formats. Likewise, it needs to have the capability to deliver IT data to OT using its preferred data protocols and formats.
While serving as the bridge between OT and IT, integration hubs generally use a graphical interface optimized for OT to link all devices in OT and connect them with all IT systems. It links protocols, converts data types, and transforms data structures. A hub also provides plugins that run custom programs to create needed manufacturing feedback loops within and across OT devices. These custom programs can leverage logic, machine learning, artificial intelligence, knowledge graphs and big data analytics from IT.
Moreover, an IIoT integration hub should be based on the “low-code” development approach to allow and empower the OT team to create its own applications and functionalities while consuming the data that is produced by the machines that they manage. This produces an integration-through-data solution that simultaneously makes the data available where it is needed, and at the same time, be dynamically processed by OT’s applications.
The end result of deploying a robust IoT integration hub is a solid foundation for an IIoT or edge-computing infrastructure that enables an organization to effectively manage and use its data in ways it could not do before. These benefits drive digital transformation of an organization to make it one that is more reliable, flexible, efficient and competitive.