The fieldbus wars from the days of yore are over. Closed fieldbuses simply don’t cut it anymore to run a modern manufacturing plant. Sure, they achieve the necessary performance, but at the expense of flexibility. Industrial Ethernet was a step in the right direction through its openness. Now that Ethernet is becoming deterministic by design, performance is no longer the deciding factor: it’s all about the data.
A Step-Change in the Right Direction
If you think back to what made fieldbuses so revolutionary, it was nothing short of a step-change: the move from analog to digital. And while replacing 25 wires with just one wire was great, the real cost savings came from less tangible benefits. Increased data capacity, high-speed networks, diagnostics: these were the features that made it possible for us to sell 60 million Profibus nodes. They led to benefits like decreased downtime, increased transparency, shorter throughputs, and overall higher efficiency.
The move from fieldbuses to Ethernet-based networks was also a step-change. Serial fieldbuses achieved their determinism by virtue of the fact that they were a closed network. Stay with me here: a control network that leverages standard unmodified Ethernet to allow deterministic data exchange to coexist plainly with other protocols all on the same physical layer is monumental. You can have your cake and eat it too, so to speak. This is what has made it possible for us to sell 25 million Profinet nodes. In hindsight though, this was an evolutionary step in the right direction; a step towards Time Sensitive Networking (TSN).
Where Does the Future Lie?
Ethernet is by definition a communication medium based on the best-effort principle. As Ethernet evolved, latency was sacrificed in favor of flexibility since its largest use-case was, and remains, the Internet. But now the IEEE is adopting and standardizing many of the concepts we’ve invented for factory automation to engineer a deterministic-by-design version of Ethernet known as TSN.
This is a fundamental shift. It means any Ethernet-based protocol can theoretically gain the performance benefits previously only associated with deterministic control networks. Any. Protocol. Think about this for a second. What does this mean for the future? All protocols can share an equal footing and the “only” differences between them are data access and feature functionality.
Using the Right Tool for the Right Task
How many Internet-based protocols can you name? HTTP, FTP, SMTP, POP, SSH, DHCP, SNMP, LDAP, and RSTP for starters. The reason we have so many is because each is suited to its purpose. The same can be said for automation networks that now share a common TSN foundation. Just because you can use a protocol for a certain task doesn’t mean you should.
The common example is Profinet vs OPC UA Pub/Sub. It doesn’t make sense to use Profinet for information exchange between office-level computers due to routability and data format. Similarly, it doesn’t make sense to use OPC UA for data exchange between field-level devices due to telegram size and overhead. OPC UA simply doesn’t have the characteristics of a modern-day fieldbus but is great for modeling and moving information. These are complementary technologies, not competing ones.
But Choose the Right Tool
Which brings us to the final point: modern fieldbuses do more than just data exchange. Such was the case with your father’s fieldbus in the 1990’s, but no longer. Today, an Industrial Ethernet protocol must be capable of motion control, data semantics, comprehensive diagnostics, functional safety, scalable media / system redundancy, energy management, process control, flexible topologies -including wireless and fiber, painless commissioning, asset management, IT integration, last-meter connectivity (IO-Link), and the list goes on.
Just remember, not all protocols situated for the same task are created equal. With performance otherwise guaranteed via TSN and more than one flavor of Industrial Ethernet on the market, differentiation comes from their capabilities, and how easy they are to use. As manufacturing complexity increases, data access and usability are the core focus for any future-proof Industrial Ethernet.
Michael Bowne
Deputy Chairman
PI
Footnote:
If you’re asking yourself, “Why TSN? This is nothing new!” The answer is: it ensures robustness even in heavily loaded networks. It is another benefit of TSN. Latency is low in lightly loaded Ethernet networks today. But as more best-effort traffic gets transmitted over the network, a near certainty in the future, TSN will ensure latency doesn’t suffer for time-sensitive (hence the name) data. E.g. this best-effort traffic might use TCP/IP, whose telegrams are relatively large.