Unified Namespace is an extended event-driven architecture, in which all data gets published regardless of whether there is an actual consumer or not. It typically leverages MQTT, or in the United Manufacturing Hub Apache Kafka as well, as a central message broker.
To understand why we are using Apache Kafka additionally to MQTT, you should check out our blog article Tools & Techniques for scalable data processing in Industrial IoT!
Benefits compared to older Industry 3.0 architectures are its scalability and ease of connectivity. In Industry 3.0 setups the general dataflow would be somewhat linear and would only flow in one direction: From the sensors to the PLC to SCADA and MES to the ERP. With a unified namespace, this would be much more connected. The idea would be that every node in the network can work as a consumer or a producer, depending on what is currently needed. A PLC receiving information from a machine learning node and then adjusts its usual algorithms would be one example of that.
In a Unified Namespace one paradigm is also fairly important: All data, whether meant to be consumed as a broker message or not should be considered important and still needs to be processed. It is meant to be seen as a business encompassing space, where every node of the business can communicate with each other.
Some more information of the person, who has coined the term “Unified Namespace”, Walker Reynolds of 4.0 Solutions:
If you have not read about event-driven architecture, we suggest that you do that.
Otherwise, you can proceed to the next chapter.