Sorry to dwell so much on data acquisition and analytics, but I can’t seem to get out of their way. It’s impacting every topic I’ve researched lately. Just as all roads lead to the Internet, cloud computing and digitalization, analytics seems to be everyone’s favorite activity once they get there. However, as with any occupation, there can be some drawbacks and pitfalls it would be good to know about and understand.
For instance, this month’s “Look closer, get granular” feature (p. 45) shows how Lubrizol is using Seeq’s software to identify tiny shifts in its batch processes. This is great news because it promises to enable more accurate and thorough optimization. However, similar to coping with large volumes of data, there can be some unexpected obstacles that crop before actual improvements are implemented and gains are achieved.
One reason we may not “see the forest for the trees” is because we’re so captivated by a new way to focus on smaller vegetation, which causes us to miss overall perspectives, strategies and procedures with greater payoffs.
However, another persistent and dangerous snag is that even useful results and valuable insights often don’t get put into practice due to various combinations of hesitation, reluctance and inertia—which are all fancy names for fear. I know I’m not the only one who knows more than a few graduate students, who never stopped researching long enough to complete their dissertations, and graduated “all but degree” (ABD). Back when I covered Chicago-area municipalities and the U.S. healthcare system, I lost count of all the blue-ribbon studies with great recommendations for improvements, which were welcomed with great fanfare, but just got shelved and never used.
Now, in this time of digitalization and rapid technical change, with network links and data sources multiplying, I worry that creeping reticence and paralysis will once again stop a lot of useful analytics from ever getting employed.
“We recently added some cool technology for adjusting the thickness of slices of pre-cooked bacon, but these devices couldn’t talk to each other until we implemented Ignition web-based SCADA software to address issues at several pain points,” says Dan Stauft, director of operational technology at SugarCreek, who spoke during a Sept. 20 panel discussion at Inductive Automation’s Ignition Community Conference in Folsom, Calif. “However, even though we found out what situations sucked, if we hadn’t made actual changes, they would have continued to suck and everyone else would have known.
“Data just by itself doesn’t do anything. It must be used and deployed. Once our plant adopted Ignition, they improved overall equipment effectiveness (OEE) by 50% in six months. But you must get the plant floor engaged, put up big dashboards, and convince the staff that the new technology won’t ‘beam them up’ or get them in trouble. Once you show people that you can add value, they’ll ask for help on other lines.”
On the reporting side, I’ve been invited to at least a couple of cybersecurity conferences, whose organizers told me I could attend, but not report on what they presented. I refused, of course, because it doesn’t help anyone if I know what to do about cybersecurity, but can’t share it. I’m not responsible for protecting any processes or plants, so what I learn about cybersecurity must be distributed to those who can act.
Of course, the solution is simply being brave enough to take useful action, especially when it’s difficult, uncomfortable or upsets antiquated mindsets and processes that many other users are likely incentivized not to change. It also means looking beyond continuing to optimize antiquated and soon-to-be obsolete technologies for increasingly inefficient and unproductive industries. I can still hear the Rolling Stones singing in “19th Nervous Breakdown,” “Your father’s still perfecting ways of making sealing wax.”
Don’t stop optimizing what's needed, of course, but keep an eye peeled for using analytics, digitalization and other opportunities to make larger, more meaningful and more epic gains.
