(www.phoenixcon.com).
He sees numerous major reasons why. One is new device-level, input/output (I/O) form factors. A second is that more vendors now deliver multiple-protocol devices. He notes that the leading Ethernet protocols in the United States—ModBus Transmission Control Protocol (TCP), Ethernet Internet Protocol (IP), Foundation Fieldbus High-speed Ethernet and ProfiNet—accommodate both process and discrete operations. Significant drops in implementation cost for an industrial Ethernet structure is also a trend.
Another trend is easier maintainability of control devices. That is possible because device-to-Web interfaces are becoming more programmable logic controller (PLC)-like, says Komarek. Another reason for easier maintainability is emerging PLC-like network-management software. That allows end-users to see the status of the entire plant-floor Ethernet system, says Komarek. Finally, another emerging development—auto-addressing—will make it easier to replace failed devices, he believes.
Early adopters
Early adopters exist, some of whom may view device-level Ethernet experimentally, says Jim Vitrano, a project engineer with Rockwell Automation Inc.’s (www.rockwellautomation.com) drives-communications group, in Mequon, Wisc. But Rockwell has a number of customers in the paper industry, as well as food and beverage industries, who are in full production using Ethernet, he says. And as controllers talk to the enterprise and simultaneously synchronize data with other factory-floor devices, Ethernet emerges as a machine-to-machine communication standard, says Rahul Kulkarni, product manager for industrial data acquisition and control at Austin, Texas-based National Instruments.
Still, end-users continue to miss the true benefit of moving Ethernet to the device level, says Benson Hougland, technical marketing director at Temecula, Calif.-based Opto 22 (www.opto22.com). By implementing the Ethernet-IP combination, the true benefit is that an end-user can run multiple protocols and solve many problems with a single technology, he says. “The beauty of Ethernet and IP, based on standard technology, is that I can have a device talk to multiple masters simultaneously.” However, if controller software cannot easily send an e-mail or talk to an Oracle database, using Ethernet-IP is not going to be better than using any other non-Ethernet based protocol, notes National Instruments’ Kulkarni.
Another benefit comes through the nature of the layered system: The physical layer of Ethernet and copper wire; IP, the next-higher layer; and then the next-higher layer of other industry-specific protocols such as ModBus TCP. “I can upgrade with new features or simply switch to other products without having to rip out hardware, networks or networking equipment,” Hougland points out. In many cases, it’s just a new firmware download, and end-users are ready to run the next generation of a protocol or new protocols, he adds.
Implementation of device-level IP technology, which allows end-users to move into wireless because it is based on IP, is the next big opportunity Hougland sees with factory-floor Ethernet. “Wireless local area networks and wired networks could be together,” he says. Rockwell Automation has drive software for diagnostics that could be connected via Ethernet directly or by wireless fidelity, notes Dave Wisniewski, a Mequon, Wisc.-based product specialist for Rockwell communications products.
To maximize benefits, automation engineers should familiarize themselves with certain tools. The first is TCP/IP addressing, Hougland says. The second comprises simple-mail-transfer protocol, simple-network-management protocol and file-transfer protocol. The third is naming, which involves IP addressing. “It allows you to sub-network. You can segment networks into traffic areas,” he says.
Kenna Amos, [email protected], is an Automation World contributing editor.
