The general manager of Knee Jerk Coatings Corp. arrives at work on Tuesday morning to a nightmare. Second and third shifts were supposed to have coated a large quantity of parts for a key customer’s important order. They coated the parts all right, but most of the parts exhibit a coatings defect that renders them unusable by the customer. A whole pile of rework.
Immediately, the GM calls the first shift to action, sorting good parts from bad ones, and processing even more raw material in an effort to fill the order due that day. The defect is also found on the parts being run by the first shift, so the rework continues to pile up.
Several managers gather at the line, throwing spaghetti at the wall. They begin by blaming the substrates without any objective data on which to base their accusations. The substrate discussion morphs into one in which suppliers are assigned fault for the problem, again without regard to whether these indictments have merit.
The group comes up with an informal list of what might resolve the problem. Try adding this. Try adding that. Change the line speed. Adjust the temperature. Rack it this way.
The GM ponders the suggestions and issues orders, selecting several items from the list of potential fixes offered by his team—a list heavily laden with tribal knowledge and black art.
Several hours later, it appears that yield is improving, though since it wasn’t initially measured, nobody knows for sure. Further, since the team changed several variables all at once, which one actually may have improved yield is anybody’s guess.
More rework. Employees are reassigned to finesse operations in an attempt to turn non-conforming product into parts the customer will accept. The more egregious defects are sent off to the stripper in order for the substrate to be recovered.
While the team frets over the mess it is creating with one customer’s part, none of the team members are focusing on anything else. Issues affecting other customers that would have otherwise been addressed in the ordinary course go unresolved, thus creating even more problems for Knee Jerk.
Employee work hours are extended with little notice in order to keep coating, sorting and finessing product. While morale is going down the sewer, labor costs are going through the roof, though everyone is too absorbed in the problem to notice.
Not until a month later, when the financial statements get published, does anyone realize what a huge financial impact this nightmare had. A lack of revenue that resulted from consuming machine capacity for rework, combined with huge labor, material and utility costs, have resulted in a financial calamity.
At Data Driven Finishing LLC the motto is, in the words of W. Edwards Deming, “In God we trust; all others must bring data.” Its second motto is “When you find yourself in a hole, the first thing to do is quit digging.” Thus, thanks to a phone call the previous evening, this company’s GM walks in with the knowledge that there was a process control issue on second shift that required the team to switch from the part it couldn’t get to run properly to one that would. There’s no pile of rework to disposition when he gets to the plant.
He pulls his team together—from the people hanging the parts to technical support—and the team defines the current state in terms of yield and process control variables.
Customer service personnel are directed to contact the customer and let him know that a process control issue is curtailing product flow. The customer is asked to define not what he wants in a perfect world, but what quantities he absolutely needs in order to maintain production. A plan is devised, with knowledge of yield and through finessing a minimal number of parts, to get the customer enough parts to keep him going until the root cause of the defect can be defined and resolved. Data Driven promises to keep the customer in the loop.
Having defined the current state and after brainstorming potential causes, the team selects the variables that are most likely causing the defect and changes one at a time, measuring the effect each change has on yield. By doing so, the team pinpoints exactly which changes have the greatest positive impact and moves quickly to make the most effective changes permanent.
By the end of the day, the problem is fixed, rework requiring stripping is minimal, and Data Driven returns to steady state.
The coatings rework nightmare. We’ve all been there. The next time it happens to you, will you be Knee Jerk or Data Driven?
