These past couple years have been busy ones for the Working Groups of Profibus & Profinet International (PI). Working Groups develop all the technical advancements made to the Profibus, Profinet and IO-Link specifications. These groups are staffed by some of the smartest engineers in the industry from companies big and small. These folks volunteer their time to help drive our technologies forward—and one Working Group has been particularly busy.
The Industrie 4.0 Working Group (I4.0 WG) was initially convened in 2015 when trends like the Industrial Internet of Things (IIoT) were first gaining traction. At the time, the group was tasked with turning concepts into reality. Instead of blindly adopting new technologies for technology’s sake, a different approach was taken. Collaborating with end users across vertical markets, the group first developed a long list of requirements. These requirements were then prioritized, which resulted in three main areas being highlighted—simplification, integration and process.
Time-Sensitive Networking (TSN) is all about making life less complicated when it comes to deterministic networking—a fundamental aspect of factory automation—over Ethernet. With Profibus, determinism was achieved via a closed fieldbus. However, since Profinet runs on standard open Ethernet, which is not deterministic by design, additional functions were built into the Profinet protocol to achieve this determinism. Namely, the protocol skips Layers 3-4 of the ISO/OSI Model and goes straight from Layer 2 to Layer 7. Furthermore, Profinet Isochronous Real Time (IRT) adds features like bandwidth reservation, synchronization and scheduling to function even faster for high-speed motion control.
Now that other industries outside of factory automation are interested in deterministic Ethernet, the features we’ve built into Profinet for 15 years are becoming IEEE standards known as TSN. This simplifies networking, as determinism can be taken for granted via TSN mechanisms. Though it might sound trivial to bring the TSN tools into the Profinet specification, it has taken a few years of continuous effort within the I4.0 WG to make this happen. In mid-2019, PI will release its Profinet@TSN spec, and we expect the first products with this technology to hit the market in 2020.
Another topic highlighted in the Working Group is horizontal/vertical integration—which focuses on reducing the effort to gain access to manufacturing line data and harmonize it for additional use. To that end, PI plans to release its Profinet/OPC UA Companion Specification by mid-2019.
Of course, machine-to-machine (horizontal) integration is nothing new for Profinet. In fact, the very first version of the protocol was used in this fashion before being expanded to I/O data exchange. With this history in mind—plus the knowledge of OPC UA’s capabilities and widespread acceptance—the I4.0 Working Group determined OPC UA to be the best way to integrate Profinet from controller to controller. In combination with TSN, OPC UA allows best-in-class machines from various vendors to communicate deterministically.
For vertical integration applications, the Profinet/OPC UA Companion Specification will begin with a mapping of Profinet diagnostics and asset management records to OPC UA objects. With such data made available via OPC UA, it can be accessed more easily by higher-level systems. For example, with diagnostics mapping, the alarms and conditions functionality of OPC UA could be leveraged by Profinet. For asset management records, the goal is to further advance the concept of the administration shell within the context of digitalization. The ability to gather detailed asset management records from Profinet devices via OPC UA will help in the creation of a digital twin for any given machine.
Whenever the topic of vertical integration is discussed, the first question always raised is about security, and this issue is being addressed by PI’s I4.0 Working Group. The challenge with industrial control system security is to strike a balance between going too far and not going far enough. With this in mind, since the first security implementations were demonstrated in April 2018, the Working Group’s initial focus has been on mitigating potential man-in-the-middle attacks.
Given that most Profinet traffic is transmitted at Layer 2 of the ISO/OSI Model, its packets are distributed according to MAC address. By definition, packets at this level do not traverse different subnets and are not routable according to IP addresses. Therefore, a malicious actor would need to beinsideyour network to do something nefarious with these packets. Of course, if a hacker is inside your network, there are bigger problems to worry about. However, because Profinet uses IP addresses in certain instances to route traffic that is not time-critical, these instances could be more susceptible to man-in-the-middle attacks. That’s why the Working Group’s efforts to secure the protocol are being focused here first.
If the TSN advancements will be felt most in discrete manufacturing, then the advancements to Ethernet known as Advanced Physical Layer (APL) will be felt most in continuous process control applications. The reasons for the difference here are that TSN mechanisms operate at Layer 2 of the ISO/OSI Model, whereas APL operates at Layer 1. The ultimate goal for the process industries is to bring Ethernet down to field-level instruments in hazardous areas.
As instruments get more complex and more data becomes available, the need for increased bandwidth to access this data becomes apparent. Ethernet provides this, with the added benefit of being a well-known standard. The forthcoming APL will be based on single-pair cabling, already familiar from the Profibus PA world. Similarly, both power and communication will be transmitted via these wires. Finally, APL will exceed the 100 m limit currently imposed on common 100Base-TX Ethernet networks.
Since this is an advance of Ethernet itself, and not just Profinet, PI is collaborating with other fieldbus organizations on this work. Therefore, the timeline for APL is a bit longer than that of the TSN implementation or the OPC UA Companion Specification. Work on APL is not expected to be completed before 2020 or 2021.
We all know that in manufacturing markets, things don’t happen quickly. But in reality, these technologies will be here before we know it.
For more information, visit PI North America at us.profinet.com.