In my last blog in this series, I discussed the OPC Unified Architecture (UA) Client-Server structure. One of the statements in that blog is “the collection of information that servers make available to clients is called the AddressSpace. The AddressSpace standardizes how objects are represented. These objects are defined in terms of variables, methods, and their relationships to other objects.” I wanted to dig in and clarify that further.
OPC UA objects and the things that comprise them – like variables and methods – are represented as nodes in the AddressSpace. Nodes are described by attributes and interconnected by references to form a graph-like model that can represent simple to complex industrial objects such as motors, valves, and pumps. OPC UA defines different classes of nodes as shown in Figure 1. Each node in a server’s AddressSpace is an instance of one of these NodeClasses. Clients access the attributes of nodes, such as the value attribute of a Variable NodeClass, by using the services provided by the server, such as the Read service.
OPC UA Information ModelsBecause OPC UA uses object-oriented techniques, it can formulate an information model that serves a specific problem domain.
By using ObjectType, VariableType, DataType, and ReferenceType NodeClasses, an information model can be constructed (see Figure 2) where the type definitions convey meaning and can represent entities found in the problem domain. This allows vendors and standards organizations to create a known object model (see Figure 3) that client applications and tools can be written against. This is one of the benefits of edge control solutions. Emerson edge control systems have an integrated object model built using OPC UA objects that allows systems – from the PLC to the SCADA system to the cloud platform – to find the right information and present that information in the right context at the right time.
Where OPC Classic provided minimal semantics such as tag name and engineering units, OPC UA provides a whole extensible infrastructure for information modeling. Leveraging this infrastructure to its fullest has the potential to transform industrial automation by providing semantic interoperability from the device to the cloud. Semantic interoperability allows client applications to generate reports, HMI screens, controls logic and analytics automatically, and when required, on demand. This is powerful because clients not only receive data identified by tag name, but information that can be interpreted so that the client can make decisions real-time on how to convey that information to facilitate better outcomes.
The next blog in this series addresses an interesting OPC UA capability – the aggregation server.