What will automation networks look like in the next five to 10 years? Wherever the Internet of Things (IoT) trend leads us, the underpinnings of future industrial networks that are universally beneficial to adopters will focus on simplification, uptime and innovation.
Using open standards in an automation network allows for manufacturers of all sizes to realize the same benefits as proprietary network solutions. Taking this a step further, what if the network were truly plug-and-play? What features would be required to enable such a network? Here are two in particular: effortless configuration and commercial off-the-shelf (COTS) hardware.
When using industrial Ethernet, effortless configuration can better be described as “allowing IT support without needing it.” For example, today a machinery maker delivers a machine with an automation solution and network configuration. The plant operator (end user) integrates the machine into his plant. This requires changes to the plant’s network configuration so that it conforms to his local IT requirements.
In the network of the future, the machinery maker will use a temporary network configuration to qualify the machine before sending it to the end user. The end user will be able to simply connect it, and the communication relationships are dynamically established at runtime. So there’s still work to be done separating the automation solution and network configuration. Within Profinet, however, we’ve already taken the step of using name-based addressing. This makes it easier for the end user to get up and running, and a big step in the right direction toward true plug-and-play.
Another way to accomplish truly plug-and-play connectivity is through COTS hardware. This is not to suggest hardware from your local electronics store should be used, only that it could be. Why? Because though an inexpensive switch made for your home or office would work just fine, it might last only 5 minutes in a production plant. Industrialized products, with the same COTS technology built in, can withstand the moisture, vibration, dust and overall harsh environment.
What exactly is that COTS technology in the network of the future? For example, already within Profinet we have technology in place to ensure the real-time transmission of production data, while coexisting on the same wire as IT data (e.g. TCP/IP, more on that later). That technology might become more mainstream with what’s known as Time Sensitive Networking (TSN).
For production data, the Profinet protocol is still required at the application layer. Its transmission, however, along with any other traffic, could be handled by TSN. TSN, as of this writing, is still very new, but has the potential to ensure real-time transmission of data like Profinet does today. If it becomes COTS technology, then this greatly clarifies the path between IT and operation technology (OT) by extending their common physical layer (Ethernet) upon which protocols like Profinet and others reside.
What good is a manufacturing environment if it isn’t up and running? In the network of the future, 0 percent downtime will be taken for granted. How do we maximize productivity by ensuring network uptime? Again, two features stand out as solutions to these issues, and are fortunately already available today: scalable redundancy and scalable security.
The most basic way to ensure uptime is at the physical level: by using a ring-shaped network. This topology creates two connections for every device. However, don’t try this at home! Doing so would cause Ethernet packets to go around in circles, eating up bandwidth and causing major headaches. Technologies like Media Redundancy Protocol (MRP) account for this by managing traffic, and in the event of a broken connection in the ring, convert the ring to a line topology.
Such recovery can take tens or a hundred milliseconds. Going further, in Profinet “bumpless redundancy,” the failover time is 0 ms. In bumpless redundancy, a sender uses two frames, going in opposite directions around the ring. So, even with a failure, a frame will still arrive at the receiver. To ensure 100 percent uptime, redundant devices and even redundant controllers can be installed, thus expanding upon simple network redundancy.
The single biggest perceived challenge to a converged network is security. In business systems, security objectives are typically ranked in terms of priority as follows:
In many automation systems, the ranking of security objectives in terms of priority is completely opposite:
To achieve scalable security, IT and OT need to agree on a security architecture that should be:
- As simple as possible but not simpler (Albert Einstein).
- As uniform as possible—if a rule is applicable in one case, it must also be followed in comparable cases.
- Understood and supported by all parties involved.
- A daily activity and not a one-time-only task.
If these can be reached, the network of the future can be a reality by allowing automation and business systems each to maintain their security objectives.
In the network of the future, OT and IT need to share the same foundation (Ethernet) while exposing their data for vertical integration. The beauty of Ethernet is it allows enterprises to use the right protocols for the right task.
There will always be different protocols used for different tasks. As network architectures continue to evolve, these protocols will begin to share the same infrastructure. That common infrastructure is standard unmodified Ethernet. So long as an automation protocol uses standard unmodified Ethernet, the benefits “come for free” as Ethernet matures.
For example, the IEEE has continually updated the 802.3 Ethernet specifications to increase the bandwidth from 10 Mbps to 100 Mbps to 1 Gbps and beyond. Profinet is one such protocol that uses standard unmodified Ethernet. As higher-bandwidth Ethernet is installed to accommodate multiple protocols, this speed increase happens automatically.
It is not just the infrastructure that needs sharing to further innovation in the network of the future; it’s also the platforms. If Ethernet is the “how,” then protocols for the data are the “what.” Profinet handles the controlling and gathering of data from devices within production systems. Via proxies, it also gathers data from non-Ethernet devices. Meanwhile, standards like OPC UA enable the communication between, within and from production systems. When used in tandem, they complete the clear path from shop floor to top floor and into the cloud where analytics can be applied.