Industrial Ethernet Protocols: Survey results show who's winning and why
The desire and need to be connected continues to grow—and gain complexity—as automation and control engineers become more heavily involved in networking at all levels. In particular, Ethernet and industrial Ethernet are finding their way onto the plant floor and into original equipment manufacturers’ (OEM) offerings with increasing frequency as related technologies evolve and ...
AT A GLANCE
The desire and need to be connected continues to grow—and gain complexity—as automation and control engineers become more heavily involved in networking at all levels. In particular, Ethernet and industrial Ethernet are finding their way onto the plant floor and into original equipment manufacturers’ (OEM) offerings with increasing frequency as related technologies evolve and improve. Ethernet has potential to simplify in the same physical layer (wires and connectors) even as many Ethernet protocols have emerged to serve varied industrial needs, sometimes within the same network.
Most engineers already use an Ethernet protocol at the monitoring or data acquisition level for SCADA and for diagnostic, testing, and maintenance applications on the plant floor. When using Ethernet or industrial Ethernet, they choose TCP/IP and EtherNet/IP most often. And, engineers are teaming with other departments, including IT, to implement systems. These conclusions are among findings of a recent survey by Control Engineering and Reed Research Group of CE subscribers’ use of and plans for Ethernet and industrial Ethernet on the plant floor and for OEM implementations.
Uses for Ethernet protocols on the plant floor are many and varied. Most common applications are SCADA and diagnostics, testing, and maintenance.
Patterns of use
In the email/online survey of 263 respondents who specify, recommend and/or buy Ethernet protocols and products, more than 19% do so for in-plant use and for OEM requirements; 51% use them only for in-plant use; and more than 30% only for OEM requirements.
About a third of survey participants have not yet formed or used network cross-functional teams. However, among those that have, the engineering staff was represented 60% of the time. Other departments frequently included on the cross-functional team include IT (30%), plant engineering and maintenance (28%), and manufacturing and plant management (20%).
Respondents were asked a variety of questions about how they use or plan to use industrial Ethernet. Most common applications for industrial Ethernet were SCADA (63%) and diagnostics,raphic for all Ethernet protocol uses.
When queried about the level they use or plan to use the protocol, most (88%) said for monitoring or data acquisition. Control applications also scored high at 84%. A significant number said they were using Ethernet at the I/O level (63%) while 60% said they were using it for manufacturing operations or execution. Other levels of use mentioned—each scoring above 40%—were for actuation, at the enterprise level, and at the sensor level. (See accompanying graphic for details.)
Major areas of Ethernet connectivity, in place or planned, include: network hardware devices (79%), PLC (78%) hardware, and PCs (77%). Of the 79% selecting network hardware devices, 84% were currently connected and 16% planned to connect in the next 12 months. Of those indicating PLC hardware, 80% said they were currently connected and 20% said they planned to connect in the next 12 months, while among those choosing PCs, 88% said they were already connected and 12% planned to do so. Other functions identified frequently as either in use or plan to use included I/O products and systems (67%), PACs (56%), variable speed drives (49%), and distributed/hybrid control systems (39%).
Choosing a protocol
The survey also asked respondents to identify which Ethernet protocols they used or planned to use for industrial applications. Among 10 Ethernet protocols represented in standards, under consideration for standards, and/or represented by organizations, TCP/IP and EtherNet/IP are far in front of all others with 73% of respondents indicating they use or plan to use one or the other. Modbus TCP also made a strong showing, with nearly half (48.5%) of all respondents saying they use or plan to use it.
Among the remaining protocols, UDP was indicated by 23.6% respondents and Profinet and High-Speed Ethernet (HSE) each by about 20%. Rounding out the list around or under 10% were SERCOS III at 11.4%, EtherCAT at 10.6%, Ethernet Powerlink at 8%, and FL-net at OPCN-2 at 4%. Slightly more than 7% said “other.” (Selection percentages appear with the protocol descriptions included at the end of this article.)
A large number of respondents (88%) are already using industrial Ethernet on the plant floor, and nearly all (99%) are somewhat satisfied or very satisfied with it. About a third started using it 2-5 years ago and nearly a quarter 6-10 years ago. Fourteen percent of all respondents have been using industrial Ethernet for more than 10 years. Of those 12% planning to start using industrial Ethernet on the plant floor, all plan to do so with the next two years.
Speed rated highest among numerous attributes desired in an industrial Ethernet protocol or Ethernet protocol used in industrial applications. Ninety-four percent of those participating in the survey said they sought “high speed” as an attribute, with 95% labeling it important and 61% calling it very important.
Other attributes that rated highly were: use of commercial, off-the-shelf Ethernet devices (noted by 86%); 10/100 Base T/TX Ethernet physical media (86%); and available devices using that protocol (85%). Among those, nearly all (99%, 94%, and 98% respectively) termed these attributes important.
Other sought-after attributes, identified by more than three-quarters of respondents, included upgradeability (79%), ability of use for control applications (79%), safety (78%), large numbers of devices use the protocol (78%), ease of data transfer to higher-level systems (77%), online training available (77%), distributed architecture (76%), wireless physical media (76%), and ability to use for monitoring applications (76%). For a complete list of attributes and their discerned importance, see this research online in the Control Engineering Resource Center at resource.controleng.com.
Respondents also reflected significant knowledge about attributes they sought in their Ethernet protocols, with 60% saying they had some familiarity with which attributes go with which protocols, and more than a quarter (27%) saying they had good working knowledge or were experts. More than half (58%) of all industrial Ethernet installations have been installed by control, plant, or operations engineers. Third-party integrators did (or will do) most of the work in 21% of these cases. Consultants did, or were slated to do, the work for 7% of the installations.
Industrial Ethernet protocols
Background follows on leading industrial Ethernet protocols in order of popularity (aided response). For more information, visit the Websites listed. For manufacturers and service providers, visit www.controleng.com/buyersguide . For system integrators, visit www.controleng.com/integrators .
Ethernet protocols are found most frequently at the monitoring or data acquisition, and control levels.
TCP/IP: Not industrial, but dominant
Network architecture of Ethernet/IP, like other CIP networks, follows the open systems interconnection (OSI) model.
TCP is the predominant transport protocol for industrial Ethernet use, noted by 73% of survey respondents as used or planned. It is the oldest protocol, defined in its most basic terms by Webopedia as “a suite of communications protocols used to connect hosts on the Internet” and “the de facto standard for transmitting data over networks.” Wikipedia says the protocol suite is a “set of layers; each layer solves a set of problems involving the transmission of data and provides a well-defined service to the upper-layer protocols based on using services from some lower layers.”
TCP/IP is also considered the dominant protocol by the Internet Engineering Task Force, a community of network designers, operators, vendors, and researchers concerned with the evolution of Internet architecture and the Internet’s smooth operation. It regularly updates the protocol and congestion control algorithms that help stabilize the Internet. IETF’s TCP Maintenance and Minor Extensions working group (TCPM WG) continues to work on maintenance (bug fixes), modest changes to the protocol, algorithms that maintain TCP’s utility, standards issues, and a TCP roadmap. The group also says it is attempting to deal with spoofed segments that can tear down connections, cause data corruption, or performance problems. IETF’s working group involved with TCP is seeking comments on the protocol on a variety of topics including TCP robustness. For more information or to comment, visit www.ietf.org/html.charters/tcpm-charter.html . For more definitions, visit www.webopdia.com and www.wikipedia.com , and Control Engineering ’s Tech Encyclopedia at www.controleng.com .
EtherNet/IP, in CIP family with DeviceNet
EtherNet/IP was noted by 73% of the respondents as a protocol they used or were planning to use. EtherNet/IP is a member of a family of networks that implements the Common Industrial Protocol (CIP) at its upper layers. CIP encompasses a comprehensive suite of messages and services for a variety of manufacturing automation applications, including control, safety, synchronization, motion, configuration, and information. Said to be a media-independent protocol supported by hundreds of vendors around the world, CIP unifies communication architecture for the manufacturing enterprise. It offers several advantages for manufacturing automation applications: Producer-consumer services let the user simultaneously control, configure, and collect data from intelligent devices over one network or use one network as a backbone for multiple distributed CIP networks. It is compatible with standard Internet protocols and standard industrial protocols for data access and exchange such as OPC. Compliance with IEEE Ethernet standards provides a choice of network interface speeds and a flexible network architecture compatible with commercially available Ethernet installation options, including copper, fiber, fiber ring, and wireless. It also offers options for industrially rated devices incorporating IP67-rated connectors (RJ45or M12) with module and network status LEDs with device labeling for ease of use. www.odva.org
Modbus/TCP widely used
Modbus TCP/IP combines a versatile, scaleable, and ubiquitous physical network (Ethernet) with a universal networking standard (TCP/IP) and a vendor-neutral data representation.
Modbus/TCP was noted by nearly half (48.5%) of respondents as an used or planned. Modbus was introduced in 1979. Modbus TCP/IP combines a versatile, scalable, and ubiquitous physical network (Ethernet) with a universal networking standard (TCP/IP) and a vendor-neutral data representation. Ethernet port 502 is assigned for the Modbus TCP/IP protocol. Modbus is media-independent with implementations over Ethernet, fiber-optic, and major wireless technologies (Wi/Fi, radio, GPRS, and Bluetooth). It gives an open, accessible network for exchange of data that is simple to implement for any device that supports TCP/IP sockets. Technically, Modbus TCP/IP takes the Modbus instruction set and wraps TCP/IP around it. With a Modbus driver and an understanding of Ethernet and TCP/IP sockets, a user can have a device up and running and talking to a PC in a few hours with minimal hardware. No licensing fees are required for using the Modbus or Modbus TCP/IP protocols, and the specification is available free and downloadable from the Modbus-IDA Website at www.modbus.org . Also available for download are implementation guides. The Website also offers access to an active, free discussion forum for users and developers. Users are not tied to one vendor for support. www.modbus-ida.org
Close to a quarter (23.6%) of survey respondents said they used or planned to use UDP. The Internet Engineering Task Force (IETF) says the UDP (user datagram protocol) makes available a datagram mode of packet-switched computer communication for an interconnected set of computer networks. UDP assumes the Internet Protocol (IP) is the underlying protocol and provides a procedure for application programs to send messages to other programs with a minimum of protocol mechanism. The protocol is transaction-oriented, and delivery and duplicate protection are not guaranteed. Applications requiring ordered, reliable delivery of streams of data should use the Transmission Control Protocol (TCP). The UDP module must be able to determine the source and destination Internet addresses and the protocol field from the Internet header. One possible UDP/IP interface would return the whole Internet datagram, including the Internet header, in response to a received operation. Such an interface would also allow the UDP to pass a full Internet datagram complete with header to the IP to send. The IP would verify certain fields for consistency and compute the Internet header checksum. Datagram congestion control protocol (dccp) is an official charter maintained by the IETF Secretariat. The working group posts information. www.ietf.org/rfc/rfc0768.txt ; www.ietf.org/html.charters/dccp-charter.html
High Speed Ethernet is a 'backbone’
HSE is well suited for use as a control backbone, supporting a range of fieldbus capabilities including complex logic functions and data-intensive process devices.
High Speed Ethernet (HSE) was tapped by about 20% of the survey respondents as used or planned. Fieldbus Foundation says High Speed Ethernet (HSE) is a key enterprise integration technology for mission-critical applications. HSE is the Fieldbus Foundation's high performance network running Ethernet and TCP/IP, and is ideally suited for use as the plant control backbone. Running at 100 Mbit/s, the technology is designed for device and subsystem integration, and enables integration of data server technology such as that provided the OPC Foundation. It supports the range of fieldbus capabilities, including standard function blocks, electronic device description language (EDDL), and application-specific flexible function blocks (FFBs) for advanced process and discrete/hybrid/batch applications. HSE supports complex logic functions for data-intensive process devices, and enables seamless information integration for plant optimization. HSE is an international standard (IEC 61158) supporting interoperability between disparate controllers and gateways from multiple suppliers. FFBs in HSE devices can be set up using programming languages such as those found in IEC 61131-3. HSE technology was designed from the ground up to support fault-tolerant networks and devices used in mission-critical monitoring and control applications. All or part of the HSE network and devices can be made redundant to achieve the level of fault tolerance required for a particular application, the group says. HSE uses standard Ethernet equipment and wiring. www.fieldbus.org
Profinet touted as all-encompassing
Among the features of Profinet is the ability to assimilate other fieldbuses such as DeviceNet, Foundation Fieldbus, and Modbus.
Approximately 20% of survey respondents said they used or were planning to use Profinet for industrial Ethernet applications. Profinet is described as an “all-encompassing industrial Ethernet” because it can be used for virtually any automation function: discrete, process, motion, peer-to-peer integration, vertical integration, safety, and others, according to Profibus Trade Organization (PTO). Because Profinet uses standard IEEE 802.3 Ethernet, it inherently works over IEEE 802.11 wireless Ethernet to permit control of untethered devices, such as automatic guided vehicles. Existing installations using wireless with Profisafe are said to be able to harness the combined power of both technologies. In motion control applications, Profinet can control 150 axes of motion in 1 ms with 1tects user investments in current plant equipment and networks by seamlessly assimilating existing Profibus installations and installations of other fieldbuses. While embracing industry standards like Ethernet, TCP/IP, XML, and OPC, Profinet is an open, industrial standard supported by PI. Founded as Profibus International, PI is reported to be the oldest and largest fieldbus organization. PI also supports Profibus, the installed base of fieldbus nodes. www.us.profinet.com
Profibus Trade Organization
SERCOS III adds capabilities to motion
FPGA-based controllers for SERCOS III are available from Xilink at www.xilinx.com.
A little more than 11% of all respondents targeted SERCOS III as a protocol in use or planned. SERCOS (SErial Realtime COmmunications System) III industrial Ethernet protocol adds new features to the original digital motion control industrial network while maintaining backward compatibility with prior versions on servo and motion-profiles, cyclic and non-cyclic communication, and the synchronization mechanism. The new version allows easy migration of existing software drivers. Compared to earlier versions, SERCOS III reduces costs through use of the Ethernet physical layer without the need for switches or hubs; the use of low-cost FPGA communications controllers; and simpler handling with Cat5e cable. It offers a higher data transmission rate (100 Mbit/sec compared to 16 Mbit/sec for SERCOS II); reduced minimum cycle time (31.25break in the ring, hot plugging for connection and removal of nodes during operation, safe communication for drive-integrated safety functions, and less cycle time. Standard (non-real time) Ethernet frames can be transmitted in parallel to the real-time channel. The controller-to-controller (C2C) synchronization and communication profile is used to interconnect motion controllers using SERCOS III. www.sercos.com
SERCOS North America
EtherCAT offers speed, flexible design
A little more than 10% of survey respondents indicated they were using or planned to use EtherCAT (Control Automation Technology). This open real-time Ethernet network (originally developed by Beckhoff) can update 1,000 distributed I/O points in 30ng expensive infrastructure components. All types of Ethernet devices can be integrated via a switch or switch port. It operates with cost-effective, standard Ethernet cards (NIC). EtherCAT technology is used to ensure compatibility and continuity throughout an I/O system in existing industrial infrastructure as well as for future system revisions and expansion. Automation software, industrial PCs, control panels, Ethernet PCI cards, and servo drives are available with built-in EtherCAT connectivity, along with a rapidly growing selection of other drives, controllers, and sensors. More than 350 participating vendors are members of the EtherCAT Technology Group. www.ethercat.org
EtherCAT Technology Group
Ethernet Powerlink is deterministic, real time
Some 8% of survey respondents say they are using or planning to use Ethernet Powerlink. The Ethernet Powerlink Standardization Group (EPSG) calls Ethernet Powerlink an ISO/OSI level 2 protocol enabling deterministic, isochronous, real-time data exchange via standard Fast Ethernet (IEEE 802.3 u). It allows high-precision data communication with cycle times as low as 100al standards, is a software-only protocol that needs no ASICs applicable with any network, and has a topology easy to learn and use. The mature technology has more than 150,000 nodes in serial machinery and is open and independent, driven by an open vendor and user group. It can be integrated seamlessly with IT networks and well-known CANopen profiles. It supports IEC 61508 safety related applications with EPL safety and is supported by more than 300 product vendors and users. www.ethernet-powerlink.org
Ethernet Powerlink Standardization Group
FL-net (OPCN-2) for controllers
Among the newest entrants, named by 4% of the respondents as used or planned, FL-net (OPCN-2) is a protocol developed by the Japan Electrical Manufacturers’ Association (JEMA). It connects factory automation (FA) controllers, such as programmable controllers (PLC) and numerical control units (NC), and enables mutual exchange of control data among controllers at high speed. Ethernet cable is used. Ethernet connects controllers with an upper layer or personal computer and is used for transmission and control of production information. An FL-net (OPCN-2) unit is implemented in a controller, such as a PLC or numerical control unit (CNC), and transmits and receives data cyclically among controllers only by assigning a station number (a node number) and common memory (also called link register) for linking in the same manner as the “CPU link unit” of a normal PLC. A particular communication program for the PLC is unnecessary. FL-net (OPCN-2) supports UDP/IP and the “FA link protocol,” located in the upper layer of UDP/IP. For more on JEMA, visit www.jema-net.or.jp/English . www.jema-net.or.jp/Japanese/hyojun/opcn_e/top-opcn.htm
June 2007 Cover Story: Who Puts the 'Industrial’ in Ethernet?
Switches, routers, hubs, gateways, connectors, and other Ethernet hardware have been industrially hardened to take what manufacturing can dish out.
Industrial Ethernet: Most important for process automation system buyers