Use Run Charts To Confirm Root Causes
A case study about reducing waste and saving costs
by Davis R. Bothe
After completing a Pareto analysis of his company's scrap costs, George, the plant manager, discovered that the biggest problem the company faced was that of oversized bores on cylinder blocks. To learn more about the problem, George quizzed Vickie, the supervisor of the department where blocks were machined. She was well aware of the situation and explained that the complexity of the operation made the process difficult for employees to master.
Unfortunately, due to ongoing layoffs, the machinist who originally ran the job was bumped to a lower paying position. Since then, a series of operators had taken over, none of whom had remained long enough to adequately learn how to run this delicate piece of equipment. According to Vickie, the personnel changes couldn't be controlled because they were a "union thing."
Vickie said that she once proposed a training center where new operators could learn how to run this equipment before actually working on the floor. But because her proposal was rejected (budget constraints), Vickie decided to display posters in that area, reminding operators that quality is one of their main responsibilities.
the root cause
Wondering if there was a way to confirm Vickie's theory that inexperienced operators were the root cause of the scrap, George asked the quality department for the scrap reports for the past several months. He then plotted the daily percentage of scrap on a run chart (see Figure 1).
George next reviewed records that tracked which employees had worked this particular job within the last three months. He overlaid the employment durations of these operators on the x axis of the run chart to determine if a relationship existed between the change in operators and the scrap rate.
This analysis clearly showed that although the scrap rate certainly varied over time, it didn't match the changes in operators. Operator A, for example, started with a low scrap rate, but his rate increased the longer he stayed on the job. If high scrap was related to a lack of operator experience, operator A certainly didn't bear this out. Operator D's scrap rate, on the other hand, began high, went low, went high again and ended low. The reverse of this pattern occurred for operator E.
Not perceiving any discernible link between scrap and operator changes, George realized that another variable must exist.
After asking some of his manufacturing engineers for ideas about what could cause bore size variation, George zeroed in on the brand of insert installed in the machine. He gathered information from the maintenance department on which brand (X or Y) was used during what time period. He then overlaid this usage rate on the run chart of scrap (see Figure 2). He noticed that brand X inserts were in the machine when the scrap rate was high, while brand Y inserts were used when the rate was low.
Testing and proving the theory
George returned to the cylinder boring operation. Learning that the scrap rate was high that day, he stopped production, removed the inserts (which were brand X as he expected) and replaced them with brand Y inserts. The scrap rate for the remainder of the day took a spectacular nosedive, as he had predicted.
The next day George had the maintenance personnel reinstall the set of brand X inserts that were removed the day before. The scrap rate for oversize bores immediately jumped, verifying the link between insert vendor and scrap. And when the use of supplier X's inserts was banned, the scrap rate for oversized bores dropped by more than 70%.
DAVIS R. BOTHE is the director of quality improvement at the International Quality Institute in Cedarburg, WI. He earned a master's degree in applied math and physics from the University of Wisconsin-Milwaukee. Bothe is an ASQ certified quality engineer and reliability engineer. He is an ASQ Fellow.