PC-based controls: Expanding plant-floor architectures from the edge to IIoT

Edge devices facilitate data processing at the plant level, increasing security and using Industrial Internet of Things (IIoT) standards.

By Eric Reiner July 10, 2017

Cloud computing is a hot topic in factory automation circles, and while Industrial Internet of Things (IIoT) technologies can offer many benefits, not every company is fully onboard with moving operations to the cloud. This could be due to security concerns, corporate policies for data access, resource availability, among other reasons. Edge computing, conversely, takes the data processing concept enabled by cloud service providers and moves it back down to the plant level, closer to the devices that create the data. This can alleviate some of the security concerns, as the data never leaves the facility, but also can serve as a stepping stone into a future cloud solution as business needs grow and change. 

Smart edge devices for data processing

"Smart" edge devices facilitate data processing at the plant level. Several tasks must be accomplished before a device can be considered a smart edge device. The first task centers on data collection from the industrial process. Once data has been acquired and stored, the edge device then accomplishes its primary task—data analysis based on preset goals or parameters. These tasks are carried out directly on the device, with the option to move the data vertically to the cloud or to other company databases for filling dashboards used by business and facilities managers.

Data can be transmitted at the edge and/or to the cloud using recognized IoT and IIoT standards, such as object linking and embedding for process control unified architecture (OPC UA), message queuing telemetry transport (MQTT), and advanced message queuing protocol (AMQP). This creates a seamless migration path for future upgrades, and PC-based control systems are best-suited for these kinds of applications because of inherent openness to IT standards for hardware, software, and networking.

Another important point to consider with PC-based control is scalability in hardware. Controls engineers can start with small processors for very basic commands and protocol translation, then migrate to powerful multi-core industrial PCs and embedded PCs for advanced data processing and analytics at the edge. The more powerful industrial PCs also can pull double duty as complete machine or line controllers.

PC-based control architectures make it possible for the same hardware platform to be deployed everywhere on the manufacturing floor. For example, this type of automation hardware can serve as the programmable logic controller (PLC), motion controller, robot controller and/or as an IoT gateway device, and much more. Thinking higher level, the hardware components used to power the cloud are typically data center-level servers run by a large IT companies, or smaller servers that are purchased by a large end-user and run on-site as a private cloud application.

Using an intelligent edge device, such as a PC-based controller, can help users filter out important data from huge masses of raw data, whether this data stays in the facility or is sent to the cloud. As part of a cloud services solution, smart edge devices can reduce the associated expenses, regardless of whether cloud service fees are based on the volume of data transmitted or the number of messages sent to the cloud.

PC-based control technology makes it possible to implement plant floor-to-cloud communication architectures that do not need expensive managed switches from a third-party IT vendor. Machine builders and manufacturers can access some pre-integrated IoT and cloud-connected technologies. The hardware involved is part of the standard industrial PC and I/O hardware offering.

All the programming or configuration to establish the IoT connectivity is conducted in the same universal platform used to program a PLC, motion control, robotics, safety, and human-machine interface. While this won’t make a motion control programmer an instant "IoT expert," it facilitates highly efficient programming among specialized programmers and teams. This way, advanced analytics and filtering can be handled at the machine in a programming environment that’s very familiar across engineering teams. 

Security for smart edge devices

The combination of a high-performance industrial Ethernet system such as EtherCAT and a vertical communication protocol such as OPC UA provides a variety of built-in security measures that do not require programming to implement. At the plant floor and machine to machine level, EtherCAT has integrated security features that block unwanted intrusion from outside sources, and it does not require the use of IP addresses. By default, EtherCAT slave devices "destroy" non-EtherCAT frames. This includes injected malware or viruses, since they are not part of the control process.

These unexpected "bad" frames of data are not forwarded by EtherCAT, so they are immediately stopped without losing important process data. EtherCAT also preserves data tunneled through standard TCP/IP devices that are connected as part of the system architecture. Since EtherCAT does not require the use of switches, there is little danger of outside frames being inserted into the control system in the first place, but if so, those frames would be destroyed automatically.

For vertical integration, OPC UA provides built-in session encryption, message signing, sequenced packets (to block replay attacks), authentication, and more. MQTT and AMQP offer similar security and encryption measures for cloud communication. Further protections can be implemented with standard IT infrastructure planning for firewalls, providing ample security for public cloud systems. If preferred, end users can consider a private cloud running entirely within the enterprise IT system. When the right technology components are in place, users can feel confident that data is encrypted and protected from intrusion.

Future smart devices must find a balance between delivering high-performance and offering a compact footprint. Despite the ultra-compact size, CPU options are available up to a quad-core processor, so this interprocess communication (IPC) can be an overall machine controller and/or an IoT gateway to push data to cloud services. IPCs can integrate industrial Ethernet protocols such as EtherCAT and cloud communication standards such as OPC UA and MQTT with ease. With the appropriate PC-based control software, the IPC hardware becomes whatever it is programmed to be. Advanced IoT and Industrie 4.0 concepts are changing by the minute, and it is this kind of flexibility and adaptability in hardware and software that will help machine builders and end-users stay ahead in the industry.

Eric Reiner is industrial PC market specialist, Beckhoff Automation. Edited by Emily Guenther, associate content manager, Control Engineering, CFE Media, eguenther@cfemedia.com.


Key Concepts

  • How smart edge devices facilitate data processing at the plant level
  • Security measures for industrial Ethernet systems and edge devices
  • How PC-based control technology benefits facilities.  

Consider this

Are built-in security features in industrial Ethernet systems and communication protocols enough to protect data processing?

ONLINE extra

Click here to view a recent webcast about edge computing.



Author Bio: Eric Reiner, IPC Product Specialist, Beckhoff Automation.