Operators of machinery and equipment continue to intensify their focus on the Industrial Internet of Things. For them to take full advantage of their IIoT-connected factories, the machinery and equipment must be able to connect to the cloud. This is achieved using what are known as edge devices.
Industrial processes are expected to achieve ever-increasing levels of effectiveness and efficiency. Energy consumption must be tracked, equipment must be quicker and easier to service, and it must be possible to measure and compare asset performance. The Industrial IoT promises all this and more with highly automated, ultra-connected machinery and production lines.
Industrial IoT solutions allow users to extract information from their machinery and equipment that extends far beyond simple alarms and event notifications. "Users can be informed when a component is showing signs of wear and when it will fail", says Ralf Pühler, Industrial IoT product manager at B&R. "Currently, only about one percent of the data a plant generates is actually utilized," he emphasizes. B&R aims to increase this rate using a modular solution package consisting of hardware and software components that can be adapted to meet customer requirements.
The modular software components of mapp Technology enable data from machinery to be collected, processed and visualized. To set up an energy monitoring system, all the automation engineer needs to do is drag and drop the mapp Energy component into their project in Automation Studio. The software component automatically collects consumption data from all of the machine's axes and then calculates and visualizes the corresponding parameters. "There are many other components in addition to mapp Energy, including one to calculate overall equipment effectiveness," adds Pühler.
Previously, machine data was only saved intermittently before being overwritten with new data. "If we want to further analyze and utilize this data, we need somewhere to store it," says Pühler. This can be a local database or a cloud-based data center.
In the age of the Industrial IoT, however, it's not just what the data can tell us about individual machines that is interesting. "I want to be able to compare machines and lines against each other, or even different production sites around the world," explains Pühler. The volume of data required for this can, in principle, be analyzed and evaluated using local computers. "However, it often makes more sense to take advantage of the virtually limitless processing and storage capacity offered by the cloud," notes the Industrial IoT specialist.
The hardware that sends the aggregated data to the cloud is known as edge devices. Edge devices? "We call them that because they are the last physical entity on the way to the cloud," explains Pühler. These devices form the interface between the operational technology (OT) at the machinery level and the IT solutions in the cloud. OT includes hardware and software components that monitor and control devices, processes and events in real time.
Data collected at the OT level can be transferred to the cloud in different ways, depending on the respective application and volume of data. "That's why we offer three types of edge devices to ensure that we have a solution to fit every situation," says Pühler.
If a sensor receives a signal once per hour, it makes practical sense to send the data directly to the cloud. Pipelines that are monitored for leaks using sensors are a good example. "Cases like that don't require real-time control, so there is no need for local control logic." Furthermore, it is sufficient for the maintenance team to check the pipeline within a few days. For simple applications such as this, a B&R bus controller suffices. This transmits unprocessed I/O signals to the cloud in encrypted form via OPC UA. "We call this option Edge Connect," says Pühler.
Where larger volumes of data are involved, it is worthwhile to aggregate the data on the machine first. This has two advantages: firstly, it reduces bandwidth requirements and costs for cloud services; secondly, it provides a sufficient buffer to prevent data from being lost in the event of a connection error. "In this case, our standard control systems can be used," notes Pühler. "We call this an Embedded Edge, and it executes real-time machine logic as well as transmitting data to the cloud."
To monitor entire production lines, data from hundreds of I/Os must be preprocessed before being sent to the cloud. In these cases, a standard controller is no longer enough. Pühler has a solution for this as well. "Here a B&R Automation PC can be combined with a comprehensive Industrial IoT platform create what we call an Edge Controller." Owing to its high processing power and storage capacity, the industrial PC can perform more advanced preprocessing and analysis than the other two edge options. Moreover, it can calculate complex algorithms such as, for example, those used in machine learning systems.
With a variety of edge architectures to choose from, it's easy to equip newly built plants for the Industrial IoT. "Edge computing isn't just for new equipment," stresses Pühler. "It also offers huge benefits for legacy equipment that until now has been operating in relative isolation." With Orange Box, B&R also has a solution for brownfield equipment. As a flexible combination of software and hardware components, the Orange Box is connected to an existing machine and can be integrated seamlessly into an edge architecture.
Protocols for a robust connection
Data is transferred from the edge to the cloud using special protocols that support the transmission of large volumes of data. B&R offers familiar queuing protocols like MQTT (Message Queue Telemetry Transport) and AMQP (Advanced Message Queuing Protocol), which allow data packets to be transferred reliably, even in cases where there the network connection is poor or intermittently unavailable. They do this by saving data packets in a queue, where necessary, to be sent at a later time.
Other protocols, including OPC UA, can be transferred over MQTT and AMQP. "OPC UA has the advantage that it is understood by all types of hardware and software in both the IT domain as well as the control system level, regardless of the manufacturer," explains Pühler. This ensures a robust connection between the machine level and the cloud, independently of which hardware is used.
Author: Carmen Klingler-Deiseroth, freelance journalist