News & Events


Would you like to subscribe?

Download full pdf

Automation Today

Latest News

Feature story

Case studies

The Fundamentals of EtherNet/IP


With the recent announcement by the ODVA of the one millionth node installation, it is fair to say that EtherNet/IP has certainly established itself. But what is it about this protocol that has seen mass adoption across so many markets? Is it because it is based on a proven worldwide standard? Is it its massive user base? Maybe it is because it offers so much more than other proprietary fieldbuses. In fact, it is a combination of all three, which has turned EtherNet/IP into a de-facto worldwide industrial standard.

Since its earliest variant, Ethernet has seen almost exponential growth, helped in no small part by the explosion into the public consciousness of the world wide web. The need for a common protocol, accessible and useable by all, was a major driver and it is fair to say that EtherNet is now the primary technology that underpins the modern information age.
One of EtherNet/IP’s primary advantages over other networks is that it is “standard, not standard based”. It uses standard hardware based on standard switches using standard management tools. Other protocols use standardised technology – but the lower down you go in their infrastructure, the less standard and more proprietary they become. Thanks to this standard architecture, EtherNet/IP makes it easy to mix commercial, business and real-time industrial controls on the same network, solving problems with off the shelf components where other solutions impose custom silicon.

Paul Brooks, European product manager for Rockwell Automation, offers an interesting analogy: “Things can, to all intents and purposes offer similar features and functionality, but in real life they are vastly different. A good example exists in the computing world. If you compare a PC to a high-end games console with regards to output – you have roughly the same thing, great graphics, great sound and virtually identical experiences. But the real issue here is the standard upon which both are based.

“The PC is rooted in an architecture that has seen constant development over the past 30 years,” he continues, “while a games console is very much proprietary and ‘closed’. Games consoles will also only be supported when they are ‘in fashion’ and the latest hottest gadget – in just a few short years though they can go from Cinderella to Ugly Sister and run out of mainstream support.”
This brings us to the crucial question: Why would a manufacturer choose EtherNet/IP? Other networks have their own strengths and weaknesses, but EtherNet/IP has enough features and functionality to make it work for a wide variety of applications. As a result, it is far more versatile than most. Because EtherNet/IP provides both control and information capabilities, it also gives companies the ability to consolidate their communications architecture.

“One of EtherNet/IP’s primary advantages over other networks is that it is “standard, not standard based.”
The primary reason manufacturers turn to EtherNet/IP is not its similarity to fieldbus, but that it offers new capabilities – many of which will be exclusive to an Ethernet-based network. These include: voice and video capability, variable topologies, IT integration, remote management, Internet connectivity and flexible security, and these are all in addition to real-time process and automation control capabilities.

The Technology
When you hear about Ethernet and/or IEEE 802.3, it is about cable, fundamental packet format and when packets can be transmitted on the network – nothing more. This standard refers to the bottom two layers of the seven-layer model. Early implementations included 10base5 and 10base2, options long since abandoned for 10baseT, 100baseT and 1000baseT, the twisted pair CAT5 and CAT6 cable found throughout offices and homes.
You will also hear about TCP/IP, or TCP/IP/UDP/IP to be technically correct. These middle layers (three and four) of the seven-layer model are the protocol used by the Internet and (likely) your business network. Using standard TCP/IP/UDP/IP layers are critical. They are used by a large number of very useful application layer protocols: http for web; ftp for file transfer; snmp for network management products; VoIP for voice and video; smtp for email; and many others.

The primary reason manufacturers turn to EtherNet/IP is not its similarity to fieldbus, but that it offers new capabilities
Ethernet topologies are also built with switches, many with special features that have been developed over many years. These special features include: Quality of Service (QoS), which enables communications to be prioritised; port mirroring, which enables users to troubleshoot problems on one switch port by connecting troubleshooting tools to another port to ‘watch’ the traffic; security features; IGMP snooping, which routes multicast traffic only to those nodes that need it; and a large number of additional features.

Finally, an industrial network needs an industrial protocol used at the very top (application) layer, sitting next to other application layer protocols such as email or web. The industrial protocol should ensure that real-time traffic is given a higher priority than other EtherNet tasks, such as servicing printers. The protocol must be able to handle the things you need to do with the industrial network, such as device configuration, data collection, peer-to-peer controller, interlocking and real-time I/O and drive control. All these are critical when deploying automation systems.
Approximately a dozen industrial Ethernet-based networks have appeared in the past 10 years, but only a few meet all of these standards. Some use standard Ethernet, but don’t comply with TCP/UDP/IP. While these networks do use standard Ethernet cable, you will need to use special/proprietary switches, and may not be able to use standard network management or troubleshooting tools.
In addition, some networks are TCP/UDP/IP ‘compatible’ but use another proprietary protocols. While some of the application layer protocols, such as email or web might work, the TCP/IP communications are sharing time with the proprietary protocol, and may not function correctly or without special configuration requirements. It is also highly unlikely that you can use standard switches, routers or firewalls with these solutions. Without compliance to all these standards, you can miss some significant benefits; indeed, your ability to mix industrial, business, and commercial traffic on the same wire or subnet may not be possible.

While a dozen or so industrial Ethernet-based networks exist, only one meets all these standards: EtherNet/IP, which uses standard, unmodified Ethernet 802.3 and TCP/IP/UDP/IP. Its application layer protocol is the Common Industrial Protocol (CIP), the same protocol used by DeviceNet and ControlNet, providing device configuration, data collection, peer-to-peer interlocking, real-time I/O, drive control and safety and motion control networking. It is perfectly compatible with other application layer protocols such as email, voice-over-IP, video-over-IP and web.