Industrial Processes and Wireless Communications – Part III: Process Automation Monitoring

The Internet is littered with articles about Industry 4.0. Adoption of digital industrial processes have accelerated since the German government led this initiative to maintain national industrial prowess. For example, with digital twins and modular flexible systems. One of the main features of Industry 4.0 is interoperability for mobile cyber-physical systems. Examples are, mobile robots, AGV(automated guided vehicles), and manufacturing infrastructure. For these processes to run over wireless infrastructure, the communications technology must deliver against strict application requirements for performance and reliability.

The details that are often unexplained in such articles, is what is actually needed from the network. I always ask the question: What are the different options to serve these application requirements? This blog is the third in this three-part series, bringing a lens on process automation monitoring, dissecting the application flow and a little detail on why the serving wireless medium must behave in a particular way to support it. We will help you understand the options open to you for digitising processes over wireless. This third blog covers the last of our 3 use-cases:

1) Mobile robots and AGV
2) Manufacturing logistics
3) Process automation monitoring

Process Automation Monitoring

Control applications automate industrial processes that operate factories, plants, and facilities, such as electric power plants, chemical plants, and oil refineries. Process automation applications continuously take measurements and use them to guide the control of the process. In process automation, there are two main areas for wireless networking; closed-loop process control (real-time), and process and asset monitoring (non-real-time).

Closed-Loop Process Control

Closed loop feedback systems, such as PID (proportional integral derivative) controllers, typically use sensors for real-time data input to the process. These sensors are designed for very high reliability (Six 9s or more availability) and accuracy, because they steer the changes in the process output. Traffic flows are usually cyclic with extremely tight cycles around 100ms.

Industrial Process and Asset Monitoring

Process and asset monitoring includes other monitoring which falls outside of the strict control loops, to capture environmental process variables such as temperature, flow, pressure, and vibration. Regular, ongoing monitoring of these variables provides information used to make forecasts, prevent downtime or equipment failure, and generally monitor the plant for security and safety reasons. Some sensors in this domain are designed to run on batteries and have a very long operational lifecycle, and therefore may also use power-saving techniques in the sensor.

Sensor data is typically more useful when combined from multiple sources together en masse, therefore the network must be able to handle a very high density, and the technology must be effective enough to enable low-cost sensors in the market, to enable the liberal implementation throughout the plant. Non-real-time solutions are often covered by various ‘industrial wireless’ standards that I tend to see on a domain-by-domain basis, and as a result operate in a siloed access network.

Traffic flows for sensors tend to be unidirectional, from the sensor to the collector, gateway, or controller. Actuators would be the other way round, whereby the controller is signalling the commands to make a change across the network. Ultimately this traffic will come back to a controller which if performing a real-time control loop, will have very stringent network requirements.

Seeking Process Continuity

For closed-loop control or very fast monitoring use-cases, the challenge with fast flow of parts and data, is that any serious anomalies whereby the process becomes out of control would mean the process has to stop. Stopping may not take much time, but the bigger issue is often restarting the process, because of the interdependency of machines, product flow, data capture and further processes must be timed correctly to restart the end-to-end industrial process. As a result, process shutdown can cause significant negative financial impact.

Instead of breaking these use-cases up in the classifications above, it’s best to look at real-time and non-real-time applications. It’s possible that a closed-loop control application may not have real-time requirements, e.g., a large tank of liquid fills up slowly but when it hits a threshold to trigger a tap actuator or pump, it doesn’t have a critical 100ms timing requirement. Whereas, object quality measurements in a fast flow of many identical items must hit a quality threshold, and therefore very fast and reliable timing is mandatory.

Requirements

Real-Time

Availability: At least Six 9s and maybe more

Determinism: High determinate, with strict timing variables to suit cyclic patterns

Reliability: All measures must be taken to bring resilience to the infrastructure

Synchronicity: Very high synchronicity must be achieved both to the communications platform and the industrial automation components

Traffic Types: Critical control-loop traffic

Non-Real-Time

Availability: Variable, and may not require a strict service level objective

Determinism: None required for basic monitoring requirements

Reliability: No mandatory redundancy

Synchronicity: None, albeit with time-stamping at collector or endpoint

Traffic Types: Regular or triggered notification messages, fire and forget

Drivers for technology decisions

Some use-cases can be solved in multiple ways, with different technologies. Secondly, technology selection is usually not the only driver for the final architecture. Especially in a brownfield environment, there may be existing technologies and processes in place that are either adopting or driving the new implementation of wireless access.

Some of these additional drivers may be:

• To use existing investments in wireless solutions
• Adhering to the technical strategy of the plant leadership team
• Optimising total cost of ownership by solving multiple use-cases with one technology
• Desire for simplicity and standardisation
• Engineering reticence to remove reliable wired systems
• Financial budgeting constraints
• Available skills and existing technical debt
• Interoperability between the industrial automation solution and the options for connecting systems wirelessly.

Cisco Solutions

I have not written these blogs to give you definitive answers on which technology is right for your use-case and how to proceed. You should make your own educated opinion on that. This is because of the unknown variables and complexity involved in every deployment scenario. Considerations must be made for tailored radio planning, brownfield systems, and specialist design to maximise the wireless performance and availability.

Each time a new architecture is selected or enhanced, a very careful analysis of all these factors and the perceived risks in available options should be assessed. Business cases should use empirical evidence and detailed design considerations where possible.

Cisco brings use-case solutions and reference architectures that enable specific business outcomes for our customers. It important to take a tailored approach each time, to ensure all aspects of your context are accounted for. We will focus on your business requirements, bringing a toolbox of options to suit your application, including Indoor, outdoor, and ruggedised WiFi, Ultra Reliable Wireless, Private 5G, LoRaWAN, and partner solutions for WirelessHART and ISA100.

Connect with your Cisco account team to discuss your industrial processes, and options for wireless connectivity in your manufacturing operations.  The Cisco account team will then help you with a deep dive on your business requirements. We will help you understand the options open to you for digitising processes over wireless.

Finally, over to you!

I’m intrigued to know if you have considered or are already using wireless communications for your robot and AGV systems. I’d love to hear your perspective and experiences.

For more context on manufacturing use cases, visit our

to see solutions created for a variety of manufacturers during their digital transformations.

Share: