All the latest news from the industry weekly compiled by the editorial team for you free of charge.
This eMail is already registered.
An unexpected error occured.
Please accept our Terms of Use.
Registration successful.
1 Rating

DIGITALISATION New industrial automation system topologies accomplished by IIoT

Aug 27, 2019

The emergence of the Industrial Internet of Things (IIoT) allows many traditional notions associated with industrial automation systems to be reconsidered. One of the traditional constructs requiring a new perspective is the topology of industrial automation systems. This article analyses the benefits of new topologies that align perfectly with inherent industrial technologies, which are made possible through the adoption of IIoT.

Starting with the introduction of digital computer technology as the delivery vehicle for process control systems in the 1960s and 1970s, industrial control system topologies have been designed for industrial operations based on the limitations and constraints of those digital technologies – not on the topology of the industrial businesses and operations. As a result, the physical and logical topologies of traditional industrial control systems have been fundamentally mismatched. Although these control systems worked and performed their functions, they were difficult to apply and maintain – at least in part because their physical and logical topologies did not naturally align with industrial operational and business architectures.

The fundamental topology and functionality of automation systems are on the verge of undergoing a significant transformation due to technological advancements, such as, big and small data analytics, cloud computing, cyber‐physical systems and the Internet of Things (IoT). The emerging automation systems will be based on natural industrial architectures (NIT) and defined by the asset architectures in the industrial plants and enterprises, not the technological constraints.

Traditional asset-centric or process-centric topological views

There are two competing perspectives for Natural Industrial Topologies – asset‐centric and process-centric. These two perspectives are based on the predominant topological view of an industrial operation. In an asset‐centric model, there are processes that are controlled within the context of assets. In process‐centric model, assets are defined as nodes within the overall processes. Traditional machine controls through PLCs have taken an asset-centric perspective from inception because early PLCs were typically aligned to a specific equipment asset. Process controls were based on a process‐centric perspective from inception. The logical topologies of traditional DCS systems were more aligned to industrial processes than the assets due to their evolution from analog process control systems. This process‐centric perspective served industry well for simple control loops up to process units. The problem with the process‐centric perspective is that it tends to significantly increase in complexity as the process extends to entire units, process areas, plants and enterprises.

Asset‐centric topologies partition the complexity of the processes into asset‐aligned components. Resolving the control strategy for the processes within an asset is much simpler than trying to resolve an overall control strategy for an entire plant. This concept is like structured analysis techniques used to simplify highly complex IT systems by partitioning them into solvable components. An asset‐centric perspective performs this partitioning naturally, making the development of control strategies much simpler.

Asset-centric Natural Industrial Topologies

Industrial businesses are designed to operate a system of industrial assets incorporating production and business processes to derive economic value. At the lowest level, plants comprise physical equipment assets – referred to herein as the primary asset of the operation. These primary assets convert materials and energy through the execution of embedded processes into valueadded production and processes along with the base sensors and actuators that are typically well documented in Piping & Instrument Diagrams (P&ID) of the operation. The primary assets are grouped into logical/physical entities called process units or work cells, which provide a functional subset of the overall production process. The units or work cells are typically further grouped into process areas that perform a larger functional subset of the overall production process. These areas are further grouped into plants. Plants are grouped into fleets. Fleets are grouped into enterprises and enterprises are grouped into value chains. This topology described is referred to as the Natural Industrial Topology (NIT) of the operation.

Another perspective that can be useful in the evaluation of industrial operations is the embedded NIT view. This view is intended to show the embedded nature of lower level assets and asset sets into the higher-level asset sets.

Asset-centric control system architectures

The key to a successful industrial operation is the effective control the processes within each asset and asset set in a manner that maximises the efficiency and profitability of the asset while simultaneously minimising the safety, environmental and security risks. To meet this need in the simplest and most effective manner, a control system should be aligned to each of the primary assets and asset sets in an industrial operation and business. The control strategies embedded in the control environment for each asset would be designed to provide real‐time asset performance control by controlling the efficiency, profitability, reliability risk, security risk, safety risk, and environmental risk to optimise the asset’s performance, leading to autonomous asset operation. The asset performance control approach would be implemented for the primary assets first. Then a unit level asset performance control strategy can be developed to coordinate the control across the primary assets comprising that unit. The unit performance controls are executed based on the understanding & trust that the asset performance for each primary asset in the unit will execute correctly. The same concept can be applied all the way up through the control hierarchy to area, plant, fleet, enterprise, and value chain. This becomes a system of systems with each subsystem performing effective asset performance control for its target asset.

For control systems based on natural industrial topologies to work effectively, a software mechanism must be developed that is designed to enable the asset performance control for lower level assets to effectively merge into the higher-level asset performance control strategy. This can be accomplished by designing each asset performance control strategy in an intelligent agent with the object-oriented characteristics of encapsulation, polymorphism, inheritance, and late binding.

Natural Industrial Topologies & the Internet of Things (IoT)

There is considerable benefit behind the concepts associated with IoT and a good match between IoT and natural industrial topologies. Industrial operations comprise a number of production assets (things) interconnected to enable the production of products. The concept of IoT is applied to industry from an asset‐centric perspective through Natural Industrial Topologies. In manufacturing and production operations, there are physical assets and asset sets as well as business assets and asset sets. Business assets and asset sets can be treated in an equivalent manner to physical assets and asset sets. Therefore, industrial businesses and value chains can be defined as a hierarchy of assets from the primary physical assets through physical asset sets that define the operation of the business through business asset sets that define the business functionality.

Cyber-Physical Systems

One concept associated with the Industry 4.0 initiative that is particularly significant to the development of Natural Industrial Topologies is that of Cyber‐Physical Systems (CPS). A CPS is a system aligned to an asset that includes the sensors and actuators necessary to provide effective control for that asset as well as the internal processes that make the asset work properly and the process control for those processes. A CPS aligned to a primary asset, or any asset set that incorporates complete asset performance control, results in an autonomous intelligent asset. In the future, equipment level CPS’ may be provided with the equipment when sold, providing significant value‐add from the equipment manufacturer.

Control functions for processes of industrial assets

Security control includes cybersecurity and any other security issues associated with the use of this asset. Effective security control should be considered an absolute prerequisite for any control architecture. By embedding security control into the intelligent agent associated with each industrial asset, the security risk of the system will decrease significantly.

The two categories of industrial risk that need real‐time control are safety risk and environmental risk. Control strategies based on the appropriate measurement of safety and environmental risks should be developed for each asset and asset set. Since environmental and safety risks present the primary constraints on driving increased efficiency and profitability from each asset, proper control of these three risks is essential for increasing efficiency and profitability. The third major area is for the control strategies associated with the ultimate objectives for the asset in the production process. The objectives here are to maximise reliability, efficiency, and profitability of the asset. There are two aspects to asset reliability that require real‐time control, the maintained state of the asset and the probability of asset failure. Both can be measured in real time at the primary asset level through deep asset knowledge. Once they are measured, the appropriate control strategy can be applied.

Benefits of control systems based on Natural Industrial Topologies

There are several obvious, and perhaps not so obvious, benefits of control systems designed to align with Natural Industrial Topologies. First, the user of these systems will inherently understand the system architecture because it is a match to the plant architecture. This will simplify the engineering task and reduce the amount of system knowledge required to effectively use the system. Also, since each node in the architecture is designed to operate autonomously in a cyber‐physical system environment, the initial configuration process will be inherently simpler because it will partition into well‐defined and natural components based on assets and asset sets. Additionally, any application or system update to a single control system associated with an asset will only impact that system. The remainder of the system should continue as set up. The single asset integrity characteristic should result in significant savings in both hardware and software, as the need for redundancy will be significantly reduced. Finally, the extended control will provide a significant improvement in the security, safety, environmental integrity, reliability, efficiency and profitability of each asset and asset set all the way up to the enterprise level.

Conclusion

Technology has progressed to a point where technological constraints no longer provide barriers to the agility needed to move toward automation systems based on Natural Industrial Topologies. These emerging systems, powered by the Internet of Things, will enable industrial companies to develop intelligent and autonomous assets across their operations that can naturally converge into overall systems of systems that will optimise the performance of each asset, asset set & entire industrial enterprises. The result will now be higher levels of industrial performance than ever before. Designing control systems to align with Natural Industrial Topologies is the first step along a path of unprecedented industrial security, safety, efficiency and profitability.

Image Gallery

  • Exponential Growth in Complexity of Process‐Centric Topologies

  • Extended real‐time control creating autonomous intelligent assets

  • Peter G Martin, PhD

    Vice President

    Business Value Consulting

    Schneider Electric

Companies related to this article
Related articles