The rise of Single Pair Ethernet in IIoT (Part 1: Interoperability)
13 April 2022
Traditionally, the networking architectures used within the industrial domain have been many and varied. With its established enterprise IT roots, Ethernet was always going to be the backbone network of choice to link the worlds of operational technology (OT) and information technology (IT) together.
However, since the arrival of Industrial IoT (IIoT), connecting all the production assets, such as sensors, actuators, robots, and the myriad of other equipment, typically involved proprietary, vendor-specific networking.
The availability of single-pair Ethernet (SPE) looks set to change this legacy approach. What is SPE technology, what are its advantages and benefits, and how can it enable cloud to sensor connectivity using a single protocol stack?
The challenges of connectivity in the industrial domain
From the shop floor to the top floor, the IIoT platform centralises real-time data from sensors, actuators, cameras, and other endpoint devices to provide manufacturing facilities actionable insight into their equipment’s performance. This information allows them to make informed based decisions to increase efficiency, optimise productivity and improve operations.
Traditionally, links to these edge-based devices use Fieldbus industrial automation protocols and rely on programmable logic controllers (PLCs) and protocol converters, or gateways, to connect to the Ethernet backbone.
Standards such as PROFIBUS DP, DeviceNet and Modbus RTU are examples of the more popular technologies. Due to the highly fragmented Fieldbus sector, many networking methodologies are proprietary, making interoperability with other implementations a significant issue. Also, they are often highly complex, requiring specialist knowledge and skilled labour to deploy, commission and maintain.
Reaping the benefits of IIoT relies heavily on using an open protocol that adheres to an industry standard to ensure interoperability between different implementations. Here, Ethernet would seem the obvious choice, and, over the years, speeds have increased to 1Gbit/s levels. But Ethernet does lack one key performance parameter, which is determinism. Deterministic networks exchange data in a precise manner with a defined latency.
Several industrial Fieldbus protocols have evolved over the years into their Ethernet TCP/IP versions. For instance, PROFINET evolved from PROFIBUS, EtherNet/IP from DeviceNet, and Modbus TCP/IP from Modbus. These Ethernet derivatives, which include others such as EtherCAT, have advanced over the years, moving from 10BASE-T (10Mbits/s) right up to 1000BASE-T1 (1Gbits/s). They offer all the benefits of commercial Ethernet but with proprietary modifications that provide lower latency and determinism.
Although most of the Fieldbus systems and Real-Time Ethernet protocols are standardised by the IEC in the 61784/61158 standards, automation devices supporting different protocols are not interoperable with each other. They often cannot coexist in a common network infrastructure. In addition, different device information models make data analysis labour-intensive and time-consuming.
However, Ethernet time-sensitive networking (TSN) has changed this by making Ethernet deterministic by design. TSN refers to a set of IEEE 802 standards that guarantee determinism and throughput in Ethernet networks, thus standardising real-time Ethernet across the industry. All of these Ethernet derivatives so far have ended at the last intelligent component of the machine because the cable and the connector’s size were too big to implement connections to the smallest sensor or other endpoint devices.
For more information about TTI, visit www.ttiieurope.com.
Contact Details and Archive...