Déjà vu Control
Standardization is a code word for reduced revenues in some quarters, and end users who champion standardization have learned to approach the topic gingerly. The phrase "same look and feel" serves as a proxy for the openness manufacturers want in controls design. While that stops short of the software compatability end users crave, it suggests a degree of standardization to wring inefficiencies out of automation projects.
"On the process side, we had engineers telling us that imposing standards was taking away their creativity, and that was a good thing," observes David A. Chappell, batch technology manager for Procter & Gamble in Cincinnati who champions standards efforts for batch processes. A member of WBF (World Batch Forum), the group that is developing the S88 framework for batch control, Chappell says, "Many of those engineers moved over to the machine side, where an artistic programming environment still exists." Unfortunately, end users like P&G pay for their artistic expression with higher integration costs and awkward reprogramming challenges that don't add value. That's why WBF is cooperating with the OMAC Packaging Work Group to develop common standards for both discreet and batch processes.
OMAC made bedfellows of Hershey Foods and M&M/Mars (now Masterfoods USA), companies more familiar with competition than cooperation. Both were early supporters of OMAC. The group has lobbied technology providers such as Rockwell Automation, Siemens and Bosch Rexroth and OEMs such as Klockner Bartelt and Schneider Packaging Equipment to support the quasi-standardization effort. OMAC's definitions of machine states for packaging have been aligned with S88's state model, and leaders of OMAC and WBF met in May to expand the effort.
For large organizations, potential savings from cooperative efforts are enormous. In a speech early this year to ARC Advisory Group's Performance Driven Manufacturing Forum, E.L. "Skip" Holmes, P&G's associate director of power, control & information systems, estimated adoption of OMAC's machine language (PackML) could save the food and consumer goods giant $15 million. As lines are reconfigured over time to run different products and package sizes, the savings from faster changeovers and less operator training to run machines with the same "look and feel" will push those savings higher.
Unilever's Andrew McDonald is another food manufacturing advocate of standardized machine controls and reusable computer code. The global manager of automation and control technology puts the potential savings for his organization at well over $100 million. That could never happen, however, if the thousands of engineers at more than 500 Unilever plants worldwide continued writing repetitive ladder logic codes that duplicate work done elsewhere and differ just enough to create integration barriers. Because Unilever has adopted the S88 standard for process batch control, "We don't need software engineers to write code anymore," says McDonald. "Wouldn't it be great to bring that down to the packaging machine level?"
Tag, you're itIt's beginning to happen, with reusable code and consistent tag names being incorporated in the controls toolboxes of automation firms and OEMs. "To be brutally honest, we ‘technology providers' operated in our own little world for years," confesses John Wenzler of Bosch Rexroth. "Ten years ago, we didn't care if our controls operated with anyone else's. Today, end users want the same look and feel in everything we do."
Tag-based names that provide standardized addresses for machines and devices in a network are an example. Traditional addresses usually were arbitrary and nonintuitive. The OMAC group addressed the issue with PackTags, nomenclature that recently was revised and improved under the direction of Larry Trunek, an engineer with Miller Brewing in Milwaukee.
"PackTags are the heart of PackML and one of the most visible, implementable projects OMAC has delivered," according to Pete Squires, a controls engineer with Schneider Packaging. "There's a little bit of Miller flavor to those tags, but Version 2 addresses shortcomings in the first version and is much more complete. Larry's effort helps a lot."
"PackTags are going to challenge the OEM who has been programming in ladder logic for 20 years, but it's going to make it easier to validate his machines, and it's going to make life a whole lot easier for the end user," adds Rexroth's Wenzler.
Rockwell Automation also backs PackTags as a solution to integrating plant-floor devices with MES systems and with each other. Built into the ControlLogix architecture is a tool called Power Programming, which lets users load standardized tag names into a controller to speed programming and simplify systems integration. Combined with the benefits of reusable code, Power Programming will slash controls installation time on a 12-axis cartoner to 170 hours from 640, Rockwell estimates.
"Power Programming supports S88, but it's way beyond batch," says Rockwell's Rick Morse, strategic marketing and business development manager. "You can take it to packaging and get the same types of information that you get at the batch end, without a systems integrator sprinkling pixy dust to optimize data flow. Major processors understand that it's the management of information that is going to give them a competitive edge. This goes well beyond the algorithms to blink the lights and spin the motors."
Standard function blocks are an essential component in the standardization effort for some end users. WBF is concentrating on part 5 of S88, which focuses on "a standard of modularization that will allow vendors to create transportable and transparent functions," according to P&G's Chappell. It would make it possible to substitute a capping machine with a Siemens controller for a unit with a Telemecanique capper, for example, with minimal pain and suffering.
That kind of interoperability was graphically illustrated at the ARC Forum in a Siemens demonstration on behalf of the Profibus user organization. Motion controllers from Siemens and Beckhoff were networked to servo drives from multiple vendors. "By hitting a button, we were able to reconfigure the hardware and go from one controller to the other without changing the PC code," reports Mike Pieper, Siemens' program manager, food & beverage. PackML provided a common interface language without requiring any knowledge of the proprietary code regulating the OEMs' drives.
The next challenge will be to apply object-based connections between machines on a production line. It requires agreement about what information has to be passed between machines, when that communication has to occur and at what speed. "A lot of details have to be ironed out, but that kind of connectivity and improved diagnostics will be good for all processors," Pieper says. "It's going to allow smaller companies to get their line efficiencies up to where they need to be to compete with the big guys."
In theory, standardization will make it easier to communicate between various programming languages, whether ladder logic, structured text, function block diagrams, sequential function charts or instruction list. Unfortunately, multiple standards groups exist, including IEC 61131-3 and PLCopen, which create some confusion. "We have two or three chefs cooking in the kitchen," Pieper concedes, but he expects the problem to resolve itself as the sponsoring groups merge. In the meantime, maintenance and support problems created when identical production lines are designed by different programmers will become a thing of the past because PLCs and motion control devices will adhere to well-understood standards.
"If we all start to use common languages like IEC 61131, everything will have the same look and feel," adds John Kowal, global marketing manager for Elau Inc., Schaumburg, IL. The controls firm uses motion function blocks to reduce the number of lines of code in a typical packaging control architecture by a factor of four.
The best solution to the machine language issue may be to render it moot, Kowal suggests. "The idea of messing around with the codes that the machine builder wrote may become obsolete," he says. "The Unilevers, Procter & Gambles and SABMillers have the engineering staffs and influence to analyze code standards and say, ‘This is the way to go.' The benefit to small and midsized processors is that they may never need to go into a machine and look at the language. Diagnostic software can be built around the de facto standard and allow end users to do 80 percent of their programming without writing any code."
While the complexity of batch processes is much greater than the discreet controls of packaging lines, processors and their suppliers see growing similarities and mutually beneficial outcomes from common standards for process and discrete machine controls. Servos are beginning to be incorporated on the process side. Some processors are ordering skid-mounted batch equipment stripped of controls for modular installation. The trend would grow if standards existed to make pre-installation of controls feasible. The goal of OMAC and WBF is to facilitate that outcome and make it possible to integrate packaging and production lines for improved efficiency and product traceability.
For more information:
John R. Wenzler, Bosch Rexroth, 847-645-3600,
John Kowal, ELAU Inc.,847-490-4270,
Rick Morse, Rockwell Automation, 440-646-3941,
Pete Squires, Schneider Packaging Equipment, 315-676-3035
Mike Pieper, Siemens, 513-236-3076,