Making things work as part of the IIoT
Sensors are on the front line of the data gathering process, which is vital for the Industrial Internet of Things (IIoT).
Reports show Industrial Internet of Things (IIoT) concepts are now starting to be adopted across manufacturing and other industries in earnest. Currently, there are around seven billion IoT devices being connected with many more expected to follow.
Technologies such as cloud infrastructure, data processing and analytics, enterprise applications, augmented reality (AR) and collaborative robots are just some of the technologies that rely heavily on the process data which is provided by smart devices. However, these smart devices – the ‘things’ of the Internet of Things (IoT) – do not always have to be new devices. It is possible to give visibility to existing legacy devices for the purposes of improving overall equipment effectiveness (OEE); real-time performance monitoring; offering a real-time warning mechanism; process analysis for predictive maintenance; or just giving the control engineer greater visibility across the plant floor. All of those things can help increase profits by avoiding downtime.
“If you want to understand the basic difference in cost and time of a system with isolated devices, and one that can collectively ‘speak’ to the systems used by the control engineer, imagine when an error occurs in a system where sensors, actuators, processor, machines, alarm systems are separate systems,” said Ivana Nikic, product marketing engineer at Moxa. “It can take a long time for an engineer to go to the local supervisory control and data acquisition (SCADA) system, gather data and then try to figure out which part of the system is causing the problem. Further, there is no easy way to predict whether a machine collected enough errors over time to fix it or replace it before the downtime happens. “So, making the ‘things’ in the system smarter – by connecting them and gathering their valuable data – is the first step towards successful IIoT solutions.”
The use of smart devices also enables machine builders to offer effective remote troubleshooting services for their end-user customers, by having remote access to their machines to analyze performance data. “Data analysis and artificial intelligence (AI) are now being used to study and adapt the manufacturing process automatically,” Nikic said.
Issues that might make companies hesitant to invest in IIoT solutions include the interoperability of the industrial communication protocols and security concerns coming from the need for information technology (IT) and Operational technology (OT) convergence. To overcome this, it is important for industry to invest in their personnel. Industrial communication equipment manufacturers can enable their devices and software to meet IIoT requirements. They can provide hardened devices and can also educate users about the different levels of cyber security protection.
Offering an example to demonstrate a successful IIoT implementation, Nikic said: “For one metal processing manufacturer with over 400 compute numerical control (CNC) machines, IIoT concepts were implemented in order to address the length of the production cycle and OEE challenges. The CNC machines and sensors were made smarter with the help of an IIoT controller, making the OT data accessible to the MES for real time monitoring and this resulted in OEE increasing to 83%. It also enabled the adaptation of IIoT for further machine and process data analysis used for predictive maintenance and process optimization.”
What’s changed for sensor requirements?
According to Richard Amery, systems sales manager at Turck Banner end users’ requirements of their sensors today are not really that much different to what has been demanded of them over the past 30 years. “End users generally do not care how the information is gathered or how it is processed. Users just want to know when component X on machine Y will need replacing or that greater efficiency can be achieved by changing from A to B at 3pm on Wednesday,” he said.
When sensors were in their infancy, they were difficult to configure. To meet industry demands sensor manufacturers needed to make their sensors very complex, in order to differentiate far smaller process differences but also needed to make them easier to set up and use. To achieve this, most sensors have been designed to gather large amounts of data and are packed with intelligence to process the data and output a simple on/off signal. Traditionally, however, this data has not left the sensor. Today this is starting to change as the value of the data is recognized.
“The data required for IIoT projects has always been available but not often utilized – because the sensor output was directly controlling the process – it is, however, now possible to connect the sensor to a programmable logic controller (PLC) input to allow its output to be recorded and time-stamped,” Amery said.
Looking to the future
Brendan O’Dowd, general manager at Analog Devices, believes the factory of tomorrow will be more agile and responsive to demands, more automated, and more reliable. It will require fewer human operators and will face less disruption due to unplanned maintenance. He puts this change down to the proliferation of miniature and high performance semiconductor sensors, alongside pervasive connectedness are creating a deluge of data on machine and process performance. “There is now more potential than ever for new applications of data analytics, such as machine health monitoring and preventive maintenance. At the same time, the increasing use of programmable hardware and software-defined electronics functions enables rapid reconfigurations of factory processes and tools,” he said.
This proliferation of sensors is generating vast flows of real-time data. Legacy communication protocols between sensor nodes and PLCs – such as 4 to 20 mA control loops – are giving way to ultrafast industrial variants of the Ethernet protocol, enabling increasing integration of OT infrastructure in the factory with IT in the enterprise.
“In response to this new demand for high speed data transfer in the factory, machine builders need to future-proof their system implementations, so that they support not only the industrial Ethernet protocols in use today, but also the emerging time-sensitive networking (TSN) variant of Ethernet, which is likely to become the standard wired networking technology for real-time industrial communications,” continued O’Dowd. “To support this transition, Analog Devices provides an Ethernet platform which enables systems to swap from one Ethernet protocol to another without the need for hardware redesign.”
On the edge
Getting the best from connected sensors and devices – the ‘things’ of the Industrial Internet of Things (IIoT), requires data from every part of the enterprise to be collected and turned into actionable information to make constant improvements. Importantly, much of the data comes from such devices that need to be computed for use in real-time to be most effective.
“Computing data in real-time requires the computing power to be placed at the ‘edge’ of the infrastructure, said Greg Hookings, business development industrial automation at Stratus Technologies. “In the near future we estimate that between 40 and 60% of generated data will be analyzed in this way and stored locally – removing the latency from cloud analytics and reducing the security implications of transferring business-critical data to and from the cloud.”
One vital consideration – and a real challenge for many industrial managers – is the need for IT skills and infrastructure to install and maintain the IT layer necessary to make the most of IIoT. “This understanding is central to the Stratus approach to computing in the industrial environment,” Hookings said. “The availability of IT proficiency at the application layer is a major pinch-point for many manufacturers. Countering this, and offering compute solutions that operate at the edge to unlock the real-time capability of IIoT at the application level, where it is often most powerful, is vital. Edge solutions will inherently improve security by reducing the amount of business critical information that is sent to and from the cloud, and this helps reduce the security concerns so often cited as a considerable hurdle to digitization.”
Looking at the obstacles businesses face looking to adopt new IIoT technologies, one of the principal concerns is undoubtedly budgeting for it. New innovations are often costlier to employ – at least until the technology becomes more readily available. However, it is important to recognize that after the initial outlay, end users will see a return on their investment in the form of improved efficiencies and extended machine lifetimes. “One cost-effective method of attaining these benefits is to retrofit smart edge computing devices to legacy machinery,” said Gavin Stoppel, product manager at Harting. “Businesses can undertake analysis at a relatively low cost by selecting a few machines at their facility for digitalization. For example, Harting’s MICA (Modular Industry Computer Architecture) contains intelligent hardware, which provides direct data processing at machine level, allowing businesses to collect and analyse their machine data.
“This information includes details such as machine operating temperatures and vibrations, which allows the engineer to undertake predictive maintenance tasks. By closely monitoring machine conditions, they can perform repairs before components break, reducing downtime and increasing system lifespan. The advent of cloud computing also enables engineers to monitor production and manage connections from anywhere in the world,” continued Stoppel.
“Live manufacturing data can also be used to improve processes, monitor the quality of products and employ flexible production methods,” Stoppel said. “By initially choosing to retrofit just a few machines, businesses can then decide when they wish to scale up the installation, allowing costs to be spread over a longer period and thorough planning to be undertaken. This approach ultimately allows businesses to focus on what they specifically want to achieve from digitalization, helping them to revolutionize their facilities and see the benefits of improved business performance.”
Data, data everywhere
With the exponential increase in the number of IoT-enabled devices, machines, and gadgets, the quantity of data to be evaluated is growing, according to Hannes Niederhauser, CEO at Kontron. “However, data acquisition with unfiltered transfer to the cloud is pointless. In Industrie 4.0 environments, reactions to sensor and actuator feedback will often be required in real time. Intelligent edge computing with compact, robust and increasingly powerful embedded boards and modules as well as industrial PCs, ensures fast, reliable and uninterrupted data processing close to the point of origin, leaving the cloud to undertake different tasks.”
With the increasing performance of boards and modules, but also with the integration of functions such as TSN, deterministic Ethernet networks become easier to set up and manage. Machines, plants and processes can be seamlessly integrated into the IT networks. “Thanks to the OPC Foundation initiative to extend OPC UA and TSN from the IT level via the controller level to the field level, this development will quickly gain momentum,” he said.
Suzanne Gill is editor, Control Engineering Europe. This article originally appeared on the Control Engineering Europe website. Edited by Chris Vavra, production editor, Control Engineering, CFE Media, email@example.com.
Original content can be found at Control Engineering.