Five Criteria for a Robust Ethernet Infrastructure

Industrial Ethernet is trending to be the principal infrastructure choice for mission-critical automation and control networks. It is typically used to connect the office with the plant floor by utilizing a single cabling platform with Ethernet connectivity and IP addressing.

This convergence of open, standards-based Ethernet communications provides all the advantages of secure, seamless interoperability in manufacturing enterprise networks -- from corporate offices to the plant floor -- and enables Internet and intra-enterprise connectivity, anytime and anywhere.

Additional benefits at the plant and enterprise level include:

• Ubiquitous access to real-time data to improve plant operations
• Real-time collaboration, inventory visibility and production planning
• Shop floor system integration with ERP for scheduling, planning, quality tracking and delivery information
• Reduced total cost of ownership (TCO) due to faster installation, less costly maintenance and upgrades

The following are five key guidelines to consider in planning and implementing a robust and reliable Ethernet infrastructure to support the industrial network.

Understand the Real Cost of Downtime

Industrial plants rely heavily on their automation, instrumentation and control data communications to relay signals between machinery, devices and control systems to activate events on an exacting and pre-determined schedule, with little or no margin for error. Plant managers and control network administrators also require optimal security and manageability so that network availability attains 99.999% uptime or better. Yet analysts report that a large percentage of unplanned downtime in industrial operations is caused by network infrastructure problems.

According to one such report, fully 72% of network faults can be attributed to failure at the OSI (Open Systems Interconnection) Layer 1 (Physical Media), Layer 2 (Data Link) and/or Layer 3 (Network). No matter what the industry, if a switch, connector or cabling system in the plant should fail, the cost of parts replacement and repair represents only a tiny fraction of the overall costs associated with production downtime.

Experience has shown that physical deterioration or electrical failure in critical data transmission components can lead to unreliable network performance and safety issues, and may lead to loss of critical data, system downtime or even catastrophic failure costing hundreds of thousands or millions of dollars. For example, an automotive assembly plant capable of producing one vehicle per minute would stand to lose profits of about $2000 to $3000 per minute for small car production, and up to $8000/min for SUV and pick-up truck production!

Evaluate the Environmental Risks

In industrial settings, environmental risks typically include physical stressors such as temperature extremes, humidity, moisture, dust/dirt, oil, solvents, and corrosive chemicals or gases. Mechanical and electrical stressors may include vigorous and repeated vibration, movement of plant floor vehicles and equipment, EMI/RFI interference, and even lightning strikes. Any or all of these can seriously degrade the components' physical integrity and electrical performance, resulting in intermittent outages or system shutdown. So it is important to plan ahead and factor environmental risks into the equation when planning and implementing an environmentally hardened industrial network infrastructure.

Specify Industrial-grade, Not COTS Components

Commercial off-the-shelf (COTS) Ethernet cables and hardware, such as those deployed in typical office environments, are not designed and engineered to withstand conditions typically present in harsh and potentially hazardous industrial settings. Industrial-grade components offer more rugged construction designed to deliver optimal performance over long service life. In fact, high quality industrial-grade Ethernet products should provide a lifespan similar to that of other automation system components -- typically 10 to 30 years, which is significantly more than COTS products can deliver.

In selecting physical media (cabling), data links, switches and hardware for the industrial Ethernet, some key factors to consider include: conformity with the Ethernet LAN.IEEE 802.3 standard; Mean Time Between Failure analysis; mounting options such as DIN rail mounted, rack- or panel-mounted, or devices that bolt securely onto machines; and a small form factor to occupy less space and allow greater density within the limited space of control panels.

In addition, depending on the installation site, users may opt to select components that provide added protection against environmental risks, such as:

• Industrial-grade cables offering extra protection in extreme environments, such as water-blocked, burial-grade, low smoke/zero halogen (LSZH), upjacketed or armored cables. A large selection of both fiber optic and copper-based cabling is available in the industrial marketplace.
• Industrial-grade connectivity components such as IP67- or IP20-rated UTP or FTP patch cords, connectors, modular jacks, plug kits, and surface-mount boxes.
• Hardware, such as switches, active network devices and accessories, should also be ruggedized for industrial environments. At a minimum, they should offer robust construction and resistance to high temperatures, vibration and EMI. There are many such products currently on the market to support both copper and optical fiber media, as well as switches capable of data speeds as high as 10 Gigabits per second.
• Conformal coatings, which can be applied to PCBs to protect sensitive electronic components in damp and/or corrosive environments.

Pre-plan for Bandwidth and Redundancy Requirements

With an ever-increasing number of Ethernet-cabled devices being added in today's automation and control networks, it is a best practice to allow for sufficient bandwidth to handle current needs, and additional headroom to accommodate future needs. This is far less costly in the long run than having to upgrade incrementally over time.

Similarly, ensuring built-in redundancy in mission-critical networks is a smart choice to prevent unplanned downtime. For a relatively small investment, redundant power, redundant data paths, and even redundant devices can help to ensure 24/7 uptime, even if a network component problem should occur.

Design for End-to-end Integration

Another best practice is to specify industrial Ethernet cabling, connectivity and hardware components from a reputable supplier capable of providing a field-proven, end-to-end Ethernet solution tailored to the application and environmental conditions. As with other kinds of information technology systems, taking a "total system" approach is typically more cost-effective and results in a more integrated system, with all products seamlessly matched and fine-tuned to deliver interoperability and consistently reliable performance.

Most industrial enterprises invest significantly in systems and gear to protect the safety and security of their workers, their equipment and their production processes. Since the industrial Ethernet infrastructure is the linchpin supporting all of the plant's automation and control functions and mission-critical information flow, it makes good business sense to select carefully and invest wisely in networking solutions designed to perform reliably and well -- day after day, year after year.

Bill Wotruba is Director, Networking and Connectivity Products for Belden, which designs and manufactures signal transmission solutions for industrial and enterprise networking. www.belden.com/industrial.