Contextualization

Introduction

Industrial sensor data, in its native form, is often available from sensors as numerical or binary data elements. The simplest industrial communication protocols provide only these basic elements, while more advanced protocols may introduce meta data such as time-stamps, quality indicators and descriptive information which enhances the value of the data. However, such simple meta data are most often not enough for applications to act upon the data the right way.

The journey from sensor data to actionable information requires that the data is set in context. Context or contextual information is any information about any entity that can be used to effectively reduce the amount of reasoning required to understand what the data means. The process of contextualization is about bringing the right context to structure-less or badly structured data. The overall objective Apis as an information gateway is to provide a standardized and unified way of accessing operational data from different assets – not only standardized on protocol, but also standardized on data semantics (context). The uniqueness of Apis lies in its ability to bring context into unstructured data, to change context of contextualized data, and to expose the same data in different contexts. Contextualization is beneficial and necessary in many scenarios, but there are two areas where the benefit of contextualization stand out; Diverse and distributed asset environments and OT (Operational Technology) Big Data scenarios.

Target Scenarios

In large organizations with distributed assets delivered by different vendors over a long period, the assets tend to expose their data on different formats and different structures. A pump from vendor A does not necessarily expose the same information as a pump from vendor B, and the pumps installed 10 years ago, display their information differently from the pumps installed today. The ability to harmonize the information structures so that the similar assets appear in a unified way is essential to be efficient when working with the data and to assure the quality of the results.

In OT Big Data applications, the contextualization stage is extremely important, since you do not always know the potential usage of the data at the time of collection, and you need to bring in context (or different contexts) to your data at a later stage. Flexibility and efficiency in the process of data contextualization is therefore a considerable advantage when it comes to handling big data in OT applications. (In IT Big Data applications, a similar concept called Schema on Read is adopted in data lake architectures. In schema on read, data is applied to a schema as it is pulled out of a stored location, rather than as it goes in.) Organizations and applications characterized by combination of diverse and distributed asset environments and OT Big Data will boost considerably by utilizing the Apis gateway.

Information models in Apis

Contextualization in Apis is built around the OPC UA standard and the information model capability of this standard. OPC UA information models are graph models structured as a full meshed network of nodes, allowing information to be connected in various ways, and expressing the semantics of the domain of interest. A key benefit of Apis as an information gateway is the capability to build and host generic OPC UA based information models. This means that OPC UA model structures can be represented in Apis, with the data variable nodes of the model hosted by Apis Hive items. This way it is possible to provide context (and different multiple contexts) to the data hosted in Apis Hive and Apis Honeystore. Information models can be constructed manually in Apis Management Studio, imported from external tools such as UA Modeler from Unified Automation or built automatically by integrating existing structure sources such as engineering databases and other master data repositories.

Asset Registry

A frequently used information model structure in industrial systems is asset hierarchies. Apis introduces the concept Asset Registry, which is based on the OPC UA ISA95 companion standard. Asset Registry can be used to organize any system, simplify engineering and enhance navigation. Utilizing the semantic modeling capabilities, Asset Registry enables system engineers to organize all enterprise and equipment data into reusable equipment (asset) types and equipment classes that can be easily instantiated for all similar assets to construct a complete system.