- ISA Member Login
 - Guest Login
 - Forgot Password?
 - Site Map
 - Join ISA
 - Sections
 - Divisions
 - Standards
 - Pressroom
 - Who's Who at ISA

Our journey begins on a train traveling across Sweden. The locomotive carries two industrial PCs, one at each end of the cabin. The screen graphics have replaced many of the indicators, annunciators, and operator controls. As the train roars down the track, an Ethernet I/O module connected to a speed sensor gathers the train's velocity information. Independent of each other, the two computers read the I/O information simultaneously, using the simple standard Ethernet connection.

The conservative nature of railroads is legendary. Why then was Ethernet selected for this application? Among the reasons is the expectancy for long service life and easy maintenance. Other reasons include the modular nature of the Ethernet system and the conclusion that Ethernet will likely remain available for a prolonged period. A reliability study highlighted Ethernet as the best choice from among the field of numerous alternatives.

Now transport yourself to an oil-drilling platform offshore in the turbulent waters of the North Sea. As part of the drilling procedure, "mud," which is a slurry of materials, is pumped down the shaft to cool the drill and carry away the debris. This constant flow of material is crucial to the drilling operation. A failure in the process can break a drill or clog a pipe, disrupting this critical and expensive investment.

Directing this mud-pumping operation to run without interruption are redundant controllers linked to the process via Ethernet and Ethernet field I/O. The Ethernet backbone's low cost was not a major factor in this application's design. If deemed necessary, a more costly solution would have been used. The IEC 1131-3 programs in the redundant controllers are synchronized over the Ethernet network, producing the deterministic performance required in this real-time system. In addition, it is far easier to protect the redundantly routed Ethernet cabling to the distributed Ethernet I/O than the large bundles of wiring classically found in the distributed control systems (DCSs) of older drilling platforms.

Much debate has centered on the suitability of Ethernet connectors and hardware in harsh industrial environments. The Ethernet controllers and I/O used on this drilling platform have been certified by the highly respected Det Norske Veritas for marine and offshore applications. This thorough testing requires compliance with 14 IEC industrial standards, including shock, vibration, and environmental conditions (e.g., dampness, heat, and performance). With its extreme demands, this application's success demonstrates Ethernet's suitability for tough environments.

The next stop on the tour takes us down under, to a major mining operation in South Australia. Here countless tons of ore are removed from the earth and processed into usable form using a sophisticated DCS with Ethernet as its backbone. This huge system consists of 420,000 I/Os reported to more than 50 computers from 148 programmable logic controllers (PLCs) located throughout this extensive complex.

It constantly monitors 63,000 alarm conditions. The computers run Microsoft Windows NT. (I have been told this is the largest Windows-based control system in the world.) The system's reliability is enhanced by several layers of redundancy. The data is managed by redundant servers that connect to the PLCs using redundant Ethernet networks, which interconnect the miles of cable spanning the entire operation. The update time on this massive system is amazingly less than 1 sec.

Those esoteric debates regarding Ethernet's suitability again rear their ugly heads when it comes to real-time control. For more than two years, this gigantic system in Australia has consistently maintained its update time well within the 1-sec specification. Of course, careful attention has been paid to the network's design, and an analysis of network loading has been performed. As with any other successful engineering project, system designers have done their homework and have designed the Ethernet network to meet the real-time performance criteria. Ethernet not only has been proved suitable for this real-time application with its large amount of data but also is probably the only network solution that satisfies this huge system's requirements.

So far, we've journeyed to areas where large or mission-critical applications involve a great deal of expensive equipment and specific requirements. Move now to a small town in upstate New York, where Ethernet I/O controls a small municipal water system. This system consists of just two well sites, a pumping station, a filtration plant, and a 1-million-gallon storage tank. Communications among these sites and the central administration facility are achieved by combining radios and phone links. The Ethernet I/O is local to the individual well sites, the pumping station, and the filtration plant.

In the above application, the I/O counts are small and time requirements slow—clearly not the reasons for selecting Ethernet. Instead, the decision was to advance to open systems architecture and avoid the problems of proprietary systems historically installed in municipal water systems. In addition, an Ethernet cable was entrenched at each of the well sites, thus ensuring the system's expandability. Now running additional cables to meet the ever-increasing demands imposed on this highly regulated industry won't be necessary. This Ethernet-based system is typical of the new water systems being installed the world over. Ethernet is fast becoming the field bus of choice in this historically conservative industry.

Focus on flexibility
Now turn your attention to a Michigan-based appliance manufacturer that maintains the quality of its millions of products manufactured each year with an Ethernet I/O system. Dozens of test cells are linked to the computers in the test lab with simple Ethernet links. The key factor in this system is flexibility. The details of tests needed change constantly. This application is included on our tour because it is typical of the trend I have observed in the implementation of testing labs, pilot plants, and other applications that require open solutions and rapid changeover. Using Ethernet I/O in these applications is not new. This particular plant installed its first Ethernet I/O in 1995.

Our world bus tour ends at an assembly plant that uses a new flexible manufacturing system consisting of modular robotic assembly stations that Ethernet I/O controls. A QNX-based controller deterministically scans the Ethernet I/O every 4 ms. I have personally studied the remarkable software that permits the rapid deployment of new assembly configurations on this modern open system. I can tell you that a PLC would not likely achieve this level of operating efficiency. The single Ethernet connection to each assembly module permits rapid reconfiguration of the system when change becomes necessary.