IIoT Answers Connected Plant Questions

Implementing a connected plant usually involves a variety of open and vendor-specific communication protocols to connect assets and then collect and transform raw data into a form that can easily be accessed and analyzed. Photo: Jens Schlueter, Getty Images News

Here are some practical IIoT technologies that will help you make your connected plant a reality.

By Rich Carpenter, Emerson

Manufacturing and processing end users have likely heard quite a bit about how digital transformation and the industrial internet of things (IIoT) are key to optimizing their operations. From a practical standpoint, how does a general term such as IIoT really qualify as a complete answer for solving any type of problem? The reality is that digital transformation is a process that combines manufacturing technology and people to provide visibility to factory or plant operations as one step toward a connected plant. The result is a foundation that optimizes operations.

Implementing a connected plant usually involves a variety of open and vendor-specific communication protocols to connect assets and then collect and transform raw data into a form that can easily be accessed and analyzed. Analytics can take multiple forms. For example, there are well-known mechanisms for analyzing energy consumption and production efficiency. Advanced analytics can also be used to generate insights through newly identified correlations across data sets, helping end users to understand equipment or process relationships that were not immediately obvious. Digital transformation is a journey and, as such, you must walk before you run. By starting at the right pace, you will gain the knowledge needed to implement the correct infrastructure.

Digital transformation is a journey of many important steps. All should be based on a robust infrastructure. Figures courtesy of Emerson

Connectivity Challenges

Traditional systems have evolved into hierarchies in which data flows from one layer to the next. Often, the truth of the data, its origin, is in the control system. This data is identified by control variables and used in logic to operate physical equipment. In addition, this data and the configuration that drives it may then be replicated into information technology (IT) and operational technology (OT) systems, and perhaps flow to a cloud infrastructure, usually for fleet or multi-plan-level analysis.

For most of these applications, accessing the data in its raw or base state, largely in the form of a tag name and an associated value, required users to re-create the information necessary to normalize the data and understand its production context. This was true for data consumers such as:

• MES (manufacturing execution systems)
• SCADA (supervisory control and data acquisition)
• historians
• analytics
• remote monitoring
• APM (asset performance management).

However, each of these data consumers often requires its own tag definitions and asset models so that data can be arranged into an appropriate context. While this was a step toward digital transformation, it introduced difficulties over time as data definitions diverged. It also required substantial effort to synchronize data sets across applications.

For today’s digital-transformation projects, these challenges are largely overcome by advancements in processing hardware, software, and communications technologies. These advancements make it easier and more practical to get the right edge data into the hands of people and systems that need it.

Many applications need edge-sourced data but have traditionally
required it in their own specialized formats, complicating digital- transformation efforts.

Connectivity Flattens Hierarchies

Old production-data hierarchies, often represented as a Purdue Model, are largely gone. For communications between various edge-located systems and devices inside a plant, OPC UA is a well-defined standard that is growing in importance to become the open protocol of choice, especially for OT applications.

Because OPC UA is platform agnostic, it works well on traditional Windows-based applications, as well as on the growing number of Linux-based IIoT applications. In addition, a new generation of edge controllers, with virtualized deterministic real-time and general-purpose operating systems on board, support OPC UA for data access.

Other fundamental capabilities built into OPC UA include:

Information modeling: Allows users to create data representations of real-world objects so that any application using the data can be informed of the context.

Discovery services: Enable applications to discover compatible sources of data on the network.

Security: Implements certificate exchanges and encryption to ensure only authorized applications can access the proper data managed by the plant OPC UA servers that may reside in equipment controllers.

It’s this set of capabilities that allows OPC UA to create plug-and-play connectivity in the typically heterogenous plant environment, provided all the players support the standard. For connectivity to IT and cloud systems, the MQTT protocol offers similar benefits.

The Module Type Package standard supports IIoT plant connectivity by enabling equipment skid functionality to be synchronized and orchestrated from a central control system.

Standardizing Equipment Definitions

Another development that promotes IIoT plant connectivity among devices of all types is the Module Type Package (MTP) standard. MTP can leverage protocols such as OPC UA but goes further to provide a standard language for defining production-equipment characteristics, capabilities, and graphical displays. It is especially useful in applications such as batch-recipe management where equipment from multiple vendors must be integrated into and orchestrated from a central controller.

By supporting the MTP standard in an equipment skid, an OEM provides a way to export the equipment software interfaces and to re-import those automatically into a centralized batch or distributed-control system (DCS). The centralized controller can then invoke the capabilities of the equipment skid in the context of the production batch it’s currently executing.

For example, a DCS system may be controlling the main process, while an equipment skid is being controlled by a programmable logic controller (PLC) or edge controller supporting MTP. Once the configurations are synchronized, the DCS system can then natively sequence the equipment through its capabilities as needed for the batch, performing operations such as mix, heat, and agitate according to the specific batch-production rules. This is an example of the tight integration possible though digital transformation.

In addition, MTP can also define the skid local-operator interface, which can be exported and then duplicated in the central control room with the system of record. MTP thus enables plug-and-play connectivity to the main process-control system and the entire equipment capabilities, control strategies, and visualization screens to be synchronized with the main plant process-control system. The MTP standard and capabilities primarily address data integration, so an OEM’s valuable equipment intellectual property remains with the OEM and is not reproduced in the main process-control system.

More Connected Benefits

OPC UA, MQTT, and MTP are examples of IIoT-enabling communication protocols and standards that support a connected plant among platforms such as a DCS, PLCs, and edge controllers. These technologies make it possible for users to create a reliable connectivity, data normalization, and persistent storage strategy, fundamental for helping plants reach stable operations while avoiding downtime surprises.

With these pieces in place, users can shift their focus to the goal of improving production efficiency. This is accomplished by implementing manufacturing intelligence, a level of analytics sitting above the equipment, where the process can be analyzed in production context.

The correct digital transformation technologies are crucial to provide a complete picture to end users. They will be informed of details for a machine, a site, and an enterprise to enable deeper understanding. End users can also consider non-machine information, such as operator actions, raw-material characteristics, and environmental conditions, to realize the full benefit of a connected plant. EP

Rich Carpenter is General Manager for Product Management for Emerson’s, St. Louis (emerson.com), machine automation solutions business and has responsibility for its portfolio of control system, operator interface, industrial PC, and industrial IOT software and hardware products for industrial automation.