WBF Presses ON!

The second day keynote was delivered by Ian Nimmo, a CONTROL contributor and expert on human factors engineering for automation and control. He talked about the same things Dave Beckman did, about how operators were not low level employees any longer, but rather highly capable knowledge workers, as were instrumentation technicians, and payscales and recognition need to change to recognise this fact. "At the end of the day," he intoned, "the process control operator has the biggest impact on production of any employee." One of the issues that it is difficult for plant managements and financial controllers to face is the need to upgrade and improve control systems on a continuous basis, Nimmo argued. "If the process 9is stable, why change? the plant management says." Which sounds suspiciously like, "The roof don't leak unless it's rainin' and when it's rainin' y'all cain't fix it nohow." There has been a lot of change in automation since the 1940s, with many external influences, including war, the role of women, the social and environmental change of the 1960s, the insistence on plant safety in the 80s and 90s, and global competition from the 90s to the present day. Nimmo told the story of this evolution at his plant, ICI Wilton, where he worked for many years. "This is no longer an isolated instrumentation system," he noted, "but rather an integrated business and production management system." Note how this is echoing Lynn Craig's comments on Sunday about having to move out of our own little world. Now, an abnormal situation is defined as one in which the operator has to intervene in the operation of the control system. The problem with this is that we've lost that hands-on interaction with the process that operators, especially in batch processing plants, used to have when running the plant on manual. What that means is that unless the operators are very well trained and supported by the control system in handling exceptions, the operator error factor can destroy all the gains made by the design of quality control systems. Many operators have poor situational awareness, and this is not helped by the common designs of HMI systems...we even have 3D piping representations, where what we really need is the ability to direct the operator's attention to what is important, and what is critical, and what is unimportant can be relegated to exception reports. One of the things we gave up in the relentless march to computer-based HMI was the natural situational awareness, the "big picture" that an old fashioned control panel gave the operators. It was possible with just a sweep of the eyes to get a complete view of the steady state operation of the plant, and using the human animal's inbuilt pattern recognition hardwiring, to be able to recognize instantly any deviation from that state. Can't do that with a computer HMI. So better designs of HMI are coming out that are lighter on cartoony graphics, and heavier on situational awareness assistance. Nimmo discussed manning theory, and showed how silly calculating number of operators required by counting loops is. Nimmo's comments were thoughtprovoking, as always, since Ian has been there, done that, and thrown away multiple t-shirts. Next, Scott Clark of Merck gave a case study on the development of a Batch Release and Material Use report for Merck. He noted that they were able to reduce the resources required to perform material accounting by four hours!!! per batch. The GUI for both reports was based on the same reporting system, so the code was the same and the interface had the same look and feel. It can also be configured for other material reports, such as waste streams, and above all, the production workers liked the system, and liked using it! Sounds like a big win.

What are your comments?

Join the discussion today. Login Here.

Comments

  • Many operators have poor situational awareness, and this is not helped by the common designs of HMI systems...we even have 3D piping representations, where what we really need is the ability to direct the operator's attention to what is important, and what is critical, and what is unimportant can be relegated to exception reports.

    One of the things we gave up in the relentless march to computer-based HMI was the natural situational awareness, the "big picture" that an old fashioned control panel gave the operators. It was possible with just a sweep of the eyes to get a complete view of the steady state operation of the plant, and using the human animal's inbuilt pattern recognition hardwiring, to be able to recognize instantly any deviation from that state.

    Interestingly, I've seen something similar in my data center over the past decade. When I was hired, the operator knew intimately what data was processing on what machine, because he or she started every update manually. Over time, the product line multiplied rapidly, launching updates was relegated to a scheduler, the update scripts emailed us log files, and we were flooded with email. Then we shunted automation-generated email from the inbox to email folders, just to keep our heads above water.

    As a result, I'm not sure that co-workers hired a year and a half ago have as much situational awareness as I had a week into my initial training. All they have to develop that awareness is the flood of email; reminders to ensure a few of the processes are where they ought to be; and alerts to exceptions and errors.

    Reply

RSS feed for comments on this page | RSS feed for all comments