🚨 Cyber Monday Sale! 🚨

Ends Midnight: Save 20% on Lean Six Sigma Training!

Unyielding Resolve: A Tale of Statistical Process Control

“Instead of asking how many reports I’ve filed, my boss is more interested in the number of parts I’ve produced.”

Recently, I was engaged by a company to review multiple plants, aiming to evaluate the effectiveness of their investment in Statistical Process Control (SPC). Sam, the plant manager, welcomed me on the first day, revealing openly, “We’re not implementing SPC here.”

Taken aback, I probed, “What’s the reason behind that?”

He responded, “Following our training, we endeavored to launch SPC, but it resulted in a decline in our output due to the extensive time dedicated to training sessions and meetings. My boss wasn’t pleased and, as a result, SPC implementation has been suspended indefinitely.”

“Is that the narrative you want included in my review?” I queried.

“Well, honestly, I don’t believe SPC brings any value to the table,” Sam shared candidly.

Not wanting to leave it at that, I asked, “Sam, what would convince you otherwise?”

His gaze wandered out the window before he returned with a smile, pointing out a man. “You see that guy? That’s Ken. Like all our workers, he’s been through SPC training. If you can persuade him of the worth of SPC, I’ll consider giving it another shot.”

“SPC will be ineffective without management backing,” I cautioned him. “Can I claim I’m representing you in this?”

Sam gave a nod of approval. “Certainly, as long as it’s reasonable.”

Ken, a robust individual in his late twenties with a mane of long hair and a thick, dark beard, didn’t look too impressed when I proposed implementing SPC in his workflow, offering my assistance in setting it up.

His gaze hardened as he replied, “My boss doesn’t care about how many reports I’ve completed. He’s only interested in the quantity of parts I’ve managed to produce.”

From his response, it was clear where Ken stood regarding SPC. I took a moment to evaluate the situation, and noticed the gauging setup to measure the parts. “What is the most crucial attribute of these parts?” I asked.

“Their thickness,” came the reply.

“Would you be alright with me checking these and handling the associated paperwork?”

Ken grunted in agreement and got back to his work.

The lapping operation in play involved two machines, each equipped with four individual rotating heads, turning both independently and around a central axis. Small rectangular components were rubbed between the rotating heads and a stationary metal bed, all the while being drenched in an abrasive slurry. This process polished the two primary surfaces of the parts to an exact thickness and smooth finish.

I concluded that samples from each head constituted a rational subgroup and charted five measurements from each head at regular intervals. Ken churned out approximately four to five batches every hour. By plotting the subgroup averages and ranges using the same scale for all charts, I was able to stay ahead of him. All of Ken’s adjustments to the machines were carefully noted.

At the end of the day, I had plotted data from 25 subgroups. Ken halted the machine, showing his first hint of curiosity.

“So, what do your charts indicate?” he questioned.

I arranged the eight charts in a way that all were visible, revealing significant variability between batches, heads, and machines.

“Look at this.” Ken indicated one of the machine’s charts. “I kept increasing the grit to maintain thickness consistency. But I guess I overdid it, causing the parts to get thinner.” His observation was accurate; one of the machines did display a significant downward trend.

The next morning, Ken was already at work when I arrived.

“I knew it!” he exclaimed, kneeling next to a large bag of grit. “I only used fresh grit on this machine. This grit is smoother and probably doesn’t lap off as much. That’s why there was no change in the thickness of parts on the other machine.” He appeared frustrated. “That means I wasted my time making those adjustments yesterday. I could’ve let both machines be, and the parts would’ve turned out better.”

Ken decided to experiment with mixing the two grits in the slurry. I consented to chart his efforts, which eventually yielded an unexpected result: All five readings from one of the eight heads were significantly higher than any previous result.

Irritated, Ken inspected the parts more closely and picked one up. “Notice how shiny this one is compared to the rest?” he asked. “That’s because it’s thicker.” He verified this with a micrometer and shook his head. “Whenever this happens, my productivity plummets.”

“Why does it occur?” I asked.

“I have no idea. It just does.”

Ken agreed to plot the charts while I consulted with the parts people. I showed Bonnie, the parts operator, the samples Ken had given me. Bonnie measured them. “They all meet the spec,” she commented.

“True, but the variance gives Ken trouble.” I explained the situation to Bonnie, who was taken aback by the news.

“We have a tolerance of ±0.003″, but we can perform much better,” she asserted.

I set up a chart and plotted a few points to demonstrate. “Wow, this is quite cool!” she exclaimed. “Is this SPC?” I nodded and promised to return at shift change to explain the chart to the night-shift operator, mentioning Sam’s name when the supervisor objected to the added work.

When I returned to Ken’s workstation, I found Melissa, an inspector, checking the parts. She was surprised to find Ken plotting charts. After studying them for a few minutes, she pointed to the outlier we’d plotted earlier in the day.

“What caused this?” she asked. Ken explained the issue. Melissa nodded and walked away. By lunch, Ken’s grit experiments had reduced the variation by approximately 25 percent. He expressed confidence in improving this further if he could use just one type of grit.

After lunch, I spoke with a grit buyer in purchasing, showing him Ken’s charts for the different grits. Normally, the buyer accepts whatever grit the distributor delivers. However, the distributor, when called, said it would send any grit the buyer requested and would also exchange any from the existing inventory.

Returning to Ken, he agreed to segregate the different brands for the exchange. He found that there was enough of his preferred grit to last until the exchange took place.

By the end of the second day, the variation was roughly half of what it had been the day before, and Ken had centered both machines close to the blueprint nominal. His productivity for the day was near a record, even considering the time spent sorting grit in the warehouse.

On the dawn of the third day, both Bonnie and Melissa eagerly awaited Ken’s arrival. Melissa had a batch of parts that were within 0.001″ in thickness. She requested Ken to run them and let her know how they performed — insisting he present her with before-and-after control charts. Ken agreed to do so. Bonnie showed him a quality control technique to sort the less consistent parts into groups of similar thickness.

The results were genuinely extraordinary. By noon, Ken had accurately identified the cycle time, slurry flow, and grit necessary to lap the parts to a specific surface finish and thickness. With few adjustments to make, he then had time to plot the control charts himself. He offered to do it before I could ask. Ken’s production for the third day set a new plant record.

I debriefed Sam before departing. He conceded that SPC was worth a second look.