Why Ethernet became the backbone of industrial communication protocols
If you have been paying any attention to the offerings of high-speed fieldbuses at all in the past 10 years, you may have noticed that most, if not all, of the mainstream offerings are based on Ethernet technology. But why is this, and why was Ethernet chosen in the first place, and what are some of the advantages of adopting Ethernet as the basis for so many different communication protocols?
In the 1990s, fieldbuses were still a relatively new and game-changing concept. Instead of bundles of individual home-run wiring that ran from field-device terminal strips or to local I/O racks, a fieldbus allowed device signals to be terminated to I/O located remotely in the field near the input/output (I/O) devices themselves, which were then interconnected with a digital fieldbus, which communicated inputs to the programmable logic controller (PLC) or other type of controller, and was sent the output data. Wiring runs were much shorter, thus saving money on cable, conduit and the associated labor costs.
In this “fieldbus war,” there were dozens of different proprietary, or near-proprietary, fieldbuses offered by many different automation companies. Many were for simple I/O, some dedicated to motion, others specifically for process control. Some were based on serial communication, others on optical or coax cable. There was not a huge incentive to be open, nor to even support third-party devices. The closed nature of most protocols was viewed as a way of keeping customers from migrating to a competitor’s control system.
Some of the main drawbacks for fieldbus users during this time resulted from the closed nature of the fieldbus systems. Dedicated and very expensive fieldbus master cards, proprietary cabling and connectors and extremely specialized diagnostic and troubleshooting tools could only be used for that particular protocol.
Around the year 2000, several automation companies and fieldbus consortiums independently began development of Ethernet-based fieldbus protocols based on the most widely used communication standard in the world—Institute of Electrical and Electronics Engineers (IEEE) 802.3, Ethernet physical layer, media access control (MAC) and other select layers of the International Organization of Standardization (ISO) Open System Interconnection (OSI) model.
So, why is this change such a big deal? Well, whether we like to acknowledge it or not, the automation industry is not that big, especially compared to the much larger IT community and consumer electronics markets.
Therefore, cabling, connectors, network cards and other infrastructure devices become less expensive because there are many suppliers making them, not just one or two. Competition drives costs down.
Another result of using IEEE 802.3 as the basis for communication is the aforementioned diagnostic tools, which were prohibitively expensive with proprietary protocols, become much more affordable or even free with Ethernet variants. Frame capture and decoding software utilities from a variety of companies makes examining and troubleshooting industrial Ethernet protocols very easy, even examining the timing and jitter of frames, command and response of bits to individual I/O channels on the network and even detecting lost, corrupt or destroyed frames.
Another advantage of utilizing a commodity technology such as Ethernet comes from who drives the innovations and improvements of the underlying standard. For the first few decades of industrial fieldbus existence, those duties and associated expense fell to the companies developing their own fieldbuses.
With Ethernet, the much larger consumer and IT community steers enhancements and developments to the communication standard. Now, those same innovations that help us stream movies, synchronize sound to video playback and play games online also help us make advances in industrial communication in factories and other industries. In this case, the innovations for consumer products help improve our industrial networking.
This does not mean we simply use the same stuff in a factory that we do in our home or office setting. The cabling, connectors and infrastructure components must still meet the requirements of the environment, the application and the industrial Ethernet protocol’s installation requirements, but these should all be in accordance with the Ethernet standards.
For instance, because factory floors are electrically noisy, shielded cable is a must, but you don’t see shielded Ethernet cable in typical office installations. And while RJ45 connectors are still perfectly fine for inside control enclosures where there is little dust, moisture or vibration, a better connector solution, such as M12 or M8 connectors, make more sense for I/O channels located outside of an enclosure, and these connectors are even compatible with washdown devices with IP67 or higher ratings. Plus, the metal connector body, being tied directly to the cable shield, connects the shield to ground without any extra installation steps.
These are outside of what you will see in typical IT installations, yet these connectors and cables are still defined as standard physical layer options for Ethernet.
So, embracing the IEEE 802.3 Ethernet standard has had enormous benefits to the industry as a whole. Not only do we now have numerous very high-performance, high-speed, low-latency fieldbuses to choose from, we also have driven the prices of the fieldbus components down. Supply chains are more forgiving when we can use standard components available from multiple suppliers, and downtime can be minimized when we can use readily available, and mostly free, diagnostic tools.
