The rise of industrial Ethernet has put more responsibility and more pressure on systems integrators when it comes to managing the network. Whereas it was once common for product vendors to set the network rules to support their specific products, the emergence of open-standard Ethernet has pushed that responsibility onto the systems integrators as they incorporate a wide range of networked products from multiple different vendors.
This has proven daunting if not overwhelming for systems integrators who previously had minimal experience managing the network. Some have even reacted by sticking with what they know, which is keeping the network at a distance.
For example, some systems integrators are using uplink/downlink connections to bring their controller onto the Ethernet network. One connection goes up into the human-machine interface (HMI) and information zone, while the other goes down into the industrial automation and control zone. This approach of providing Ethernet through the backplane serves the same function as a managed switch, but only delivers a fraction of the speed that a managed switch and standard IT Ethernet technology offers.
As a result, the full power of Ethernet technology is not embraced.
Avoiding the network only does a disservice to systems integrators and their customers. And while nothing can replace training and professional development to help systems integrators become better experts in network architectures, there are some key considerations that can be applied to almost every industrial Ethernet project.
Respect the Spec
Control infrastructures are specified. Network infrastructures should be, too.
Systems integrators will often install the network infrastructure without using a set of documented specifications. But a physical network and its many components involve a diverse set of requirements, including varying pathway distances, multiple types of equipment used in different locations, and varying enclosure requirements.
Deploying a physical network infrastructure without first establishing specifications can make installation more challenging and costly. It can also lead to unexpected downtime events and expensive retrofits in the future.
Make It Redundant
Also like a control system, networks should be built for redundancy through the use of resilient topologies. A Redundant star topology uses redundant switches to maintain network uptime, even in the event of multiple points of failure, while a device-level ring (DLR) topology is a single-fault tolerant approach that offers simplified cabling and doesn’t require an additional switch.
In a DLR topology, proper pathway planning can be critical. Laying two connections that provide single-fault redundancy in the same pathway, for example, will make the topology fruitless if a single incident can take them both out at once. Instead, use diverse pathway planning.
Additionally, consider using color-coded or lock-in connectors when deploying these topologies to help eliminate confusion or prevent incorrect port connections.
Structured Vs. Point to Point
A point-to-point cabling approach, in which devices and machine ports are directly connected, is unplanned and deployed on an as-needed basis. It should only be used in applications that have a low cable density, in environments that are unlikely to have large or multiple changes occurring in the future, and in areas where it is difficult or impractical to use cable-connector interfaces, such as patch panels.
But networks are growing in size and complexity – IHS Global and IMS Research estimate 160,000 new industrial Ethernet nodes are connected every day – and structured cabling is becoming the more sensible approach. Adopted from the IT sector and governed under the ANSI/TIA-568-C series of standards, structured cabling is a planned cabling method used to manage industrial connections and local Ethernet ports.
Structured cabling is ideal for high cable-density environments, as well as for applications that require multiple points of connectivity or that will likely experience multiple changes in the future. In addition to improved cable organization and reduced network sprawl, structured cabling can help reduce the risk of incorrect terminations. Remember: It only takes one mismatch or cross patch to render a network inoperable.
Design for the Future
It’s tempting to design a system for today to address the most pressing near-term needs while also controlling costs. But that approach can actually be more expensive in the long run if the network can’t keep pace with operations.
An industrial Ethernet network infrastructure that can scale up with a plant’s long-term growth and adjust to changing operational needs will help make future network changes easier, less disruptive and less costly. Key considerations for a future-ready network include designing in a higher bandwidth to meet future data demands and choosing a more robust switching technology to accommodate future connections.
Know Your Environments
Network equipment can be exposed to a wide range of environmental factors on the plant floor, including extreme temperatures, humidity, shock, vibration and chemical exposure. Cables and wires must be ruggedized and appropriately protected against these conditions.
The ISO/IEC 24702 and TIA-1005 standards recommend using the MICE (Mechanical, Ingress Rating, Climatic, Electromagnetic) classification system. It categorizes multiple environmental classes to help select materials that are suited for different conditions. Commercial-grade physical infrastructure may be appropriate for an area such as a control room where the MICE system shows no hazards, while harsher conditions that do present hazards will drive higher requirements, such as using shielded versus non-shielded cabling.
Greater connectivity and availability of valuable performance data naturally requires a strong commitment to security. Defense-in-Depth (DiD) security is an industry-recommended best practice that establishes multiple layers of protections against network threats.
From a physical security perspective, this can include blocking ports to help limit personnel access to the network and prevent unauthorized device connections. Color-coded patch panels can help reduce inadvertent patching mistakes that risk bypassing the DMZ firewall between industrial and information zones. Keyed and lock-in jacks can also better control connectivity to certain switches.
Why Go It Alone?
Industrial networks have never been more complex or important than they are today, from control systems that hinge on the network to the growing reliance on plant-performance data for business decisions. As a result, ensuring that an Ethernet network’s deployment is smooth and timely is critical.
Solutions providers offer expertise and services that systems integrators may not have when it comes to network design and implementation. Their project-management support capabilities and implementation services can help systems integrators meet aggressive timelines, avoid delays and resolve implementation issues as they arise.
To learn more, sign up for the Industrial IP Advantage industrial network design training here.