Predicting Maintenance
by Supriya Nayalkar
Downtime used to be a daily occurrence in The San Diego Union-Tribune's post-press department. Hoppers on the inserting machines used to bind up regularly, and the gripper chain would churn out rejects incorrectly. Each time, the paper's electrical crew would toil over the problem for as long as an hour.
Not any more.
Six months ago, Naveen Sharma, the paper's electrical supervisor, devised an innovative, time-saving system to gauge the health of the paper's equipment and anticipate any mechanical restrictions and potential electrical failures.
First, he counted the number of hoppers on the inserting machines, then unsynchronized the gripper panel and took independent readings of speed, voltage, current and load to provide benchmarks for the equipment's optimal working conditions.
Sharma then set up a Microsoft Access database on his Windows PC, and entered the benchmark data along with readings crew members collected once a week. Doing so allows Sharma to watch for changes in performance.
"Any variation in trends in a graph lets me know of potential electro-mechanical failures, which I look at as soon as possible," he says. For instance, "variations in the current of the gripper chain tell me whether the chain is too loose or too tight." Sharma says he hasn't quantified the performance of the system, in place for less than a year, but notes that downtime went from being a daily problem to a rare one.
The database proves useful for both preventive and predictive maintenance. And it's a "good training tool for new hires, who can look at the database, see what the symptoms are and what was done to it before, and come up with a better diagnosis."
While workers view the Access data through the newspaper's main server, management evaluates equipment performance using a supervisory program run in Microsoft's Visual Basic. An extension of the Basic programming language, Visual Basic lets users analyze data and reports through querying.
"Queries let you find out whatever you want to know about the machines, such as how they compare to other equipment or how to zoom in on which machine a problem is related to," says Sharma.
The database approach has proven so effective in minimizing downtime, The Union-Tribune now applies it to tackling pressroom problems as well.
Supriya Nayalkar is a free-lance writer based in Washington, D.C. E-mail, SupriyaN@aol.com.