Tales from the Trenches – 2: Assignable Cause and Corrective Action Codes

Tales from the Trenches is an ongoing series of blogs and videos designed to help manufacturing quality professionals deal with the issues that arise on the plant floor...no matter what industry you are in.
 
As I mentioned in the first article in this blog series---about reducing customer complaints—if the quality of your products is consistent and kept to a high standard, then there is really no reason for your customers to look elsewhere. And the surest way to ensure that your manufacturing operations are under control and kept in check, which leads to consistent, high-quality products, is to seek continuous process improvement.
 
With this in mind, in this “Tales from the Trenches” blog article, I’m turning my focus to how I have used Assignable Cause and Corrective Action codes to collect useful information for process improvement. In this article, I’ll also review Pareto charts and how I’ve used them during my tenure in the food and manufacturing industries to focus our continuous improvement projects to gain the most value from our work.
 
And I’ll show you how I used InfinityQS statistical process control (SPC) software to reduce defects and rework, lowering product costs.
Lowering Product Costs through Quality Control

Meaningful Process Improvements Require Solid Data

It’s widely understood that relevant, accurate data is a requirement for a solid quality assurance system. That data provides the knowledge necessary to recognize areas for improvement and to quantify the success of your organization’s continuous improvement activities.
 
Nobody wants to put effort into a project that cannot be proven to be worth the expense. I know from first-hand experience how frustrating it can be to make improvements when the baseline data and improved state cannot be evaluated in a meaningful way. It’s frustrating for everyone involved.
 
This is something I have seen time and time again in the food industry. But, for every unclear result, there are many examples of successful projects that utilize existing data collection methods to prove that improvements did, in fact, occur.
Quality Control in the Food Industry

For Example

The following is a brief look at a problem I encountered with heat sealing equipment, as well as the data I was able to collect using InfinityQS ProFicient software, which helped me solve the problem.
 
We have all seen consumer goods in plastic containers with a foil lid. Think yogurt, or nuts, or a can of potato chips. These products all employ a similar technology using heat, time, and pressure to affix a foil lid to a container, preserving the freshness of the product until it reaches the consumer.
 
When the sealing equipment is working well, things work great; but, in my experience, these are finicky machines that don’t handle variation well at all.  Whether it’s a change in foil lots, an unplanned downtime event, or any one of dozens of other variables, something is always causing poor seals, requiring operator and maintenance resources to bring the system back into control.
 
Besides the frustrations at the operator level, poor seals lead to rework, scrapped packaging materials, and of course, customer complaints. In my case, the customer complaints were often for seals that were too strong, sometimes impossible to open without damaging the container.
Sealing Machinery in the Food Industry

The Fix is In

Obviously, something had to be done, but where to start? Each operator and mechanic had their own theory about why the machine was breaking down and their own unique way to correct the issue. Because of this, there was no consensus on how to improve the process. What I needed was some collated, curated data. What I needed was a good Pareto chart.
 

Pareto Charts

Vilfredo ParetoWhen trying to find the root cause of a problem, or the appropriate corrective action for your solution, a Pareto chart is an indispensable tool. Applied to quality assurance by Dr. Joseph Juran, the Pareto principle is named in honor of Italian economist Vilfredo Pareto, who developed the idea to illustrate that 80% of the land in Italy in the late 1800s was owned by approximately 20% of the population.

This idea of an “80/20” rule has become commonplace in modern times and generally posits that 80% of outcomes result from only 20% of the potential causes.
Pareto's Law 
The Pareto chart (shown below) is simply a combination of a bar chart and a line chart, but don’t let the simplicity fool you—it’s a powerful tool.
Pareto Chart
Whereas many bar charts are sorted by time, date, alphabetical order, or some other categorization, a Pareto chart’s bars are ordered by their value, with the tallest (longest) bars coming at the left or top and decreasing right or down. This order enables you to quickly see which categories make up the “vital few” in whatever you’re measuring.
 
The line graph helps to highlight how “vital” those few are, by acting as a running total for the graph. In a typical Pareto chart, where the 80/20 rule is evident, the line graph appears to take on a logarithmic or asymptotic (a curve and a line that get closer, but do not intersect) shape—rising quickly then leveling off as it nears 100%. This reinforces the concept that most of the “activity” in the chart is typically in the first few categories.
 
When your data is plotted on a Pareto chart, the biggest bang for your buck happens when you first address the issues in those tallest (or longest) bars.
Chart Analysis on the Plant Floor

Operator Input

Inspecting product packaging seals is not something that, at first glance, seems like a good fit for SPC tracking. Since it is a pass/fail process and each package is inspected, end-user complaints are both unlikely and rare.
 
The data gathered could generally inform you of a potential problem (or that whatever solution had been implemented was successful), but that’s about it. In the case of the defective seals, notes generated by operators (the hallmark of any good SPC program) were sporadic and unstructured, making them fairly useless. And a Pareto chart of that data was of no value.
 
However, improving the structure of operator notes was important—so, the Assignable Cause Code and Corrective Action Code functionality of ProFicient was worth its weight in gold. Leveraging the rules and structure of those components yielded data that would have made Vilfredo Pareto and Joseph Juran proud. More importantly, it yielded the type of data that could be used to quantify the improvements we were making to the sealing process, confirming that our focus on operator adjustments was the key to improving the sealing process.
Keep Operators Happy 
To get the data I needed, I first interviewed operators and mechanics to see what they thought were the root causes and best solutions to each seal problem. This was the most time-consuming part of the process, but I considered it critical that a majority of affected employees were interviewed to get a full set of scenarios.
 
Using that data, I created a set of Assignable Cause codes and a subsequent set of Corrective Action codes. After all the codes were created and communicated to the team, I enabled the alarm notification rules to require an Assignable Cause and Corrective Action for each failed seal inspection.
ProFicient Alarm Notification

Bring On Assignable Cause and Corrective Action Codes

I was soon rewarded for my efforts with a stream of useful, quantitative data. When a baseline of data was established, we set to work identifying the 20% of problems that were causing the 80% of our defects.
 
By using the Assignable Cause/Corrective Action feature in ProFicient, I was able to make it quick and easy for operators to enter their codes, keeping them happy both by the lack of extra documentation and also in the knowledge that they were actively participating in a project that was going to improve their work by reducing rework and improving customer satisfaction.
Selecting Assignable Action Codes  Selecting Corrective Action Codes

Keep Your Operators Happy

One thing we kept in mind while implementing our new, structured Assignable Cause and Corrective Action codes was operator satisfaction. I had to be mindful that these codes must not be seen as adding barriers or restrictions to their work. It’s important that operators be respected as the process experts they are, but it’s also important to have some “guardrails” on the data they record.
 
The key to balancing these two forces is to have a high level of operator input and two-way communication around the implementation.
  • Everyone needs to know the goals, timing, and purpose of the work.
  • Even more than that, operators need to see their inputs showing up in the final product.
Taking all those items into consideration led to a successful implementation of the Assignable Cause and Corrective Action code functionality.
 
In the end, the root cause of our seal issue was the idea that if a strong product seal is good, then a VERY strong product seal must be better. Our seal inspection process focused on identifying poor seals but wasn’t very good at finding seals that were too strong and difficult to open, and those were the seals causing customer dissatisfaction.
 
Using the data, we quickly found that the most common corrective action when a seal issue was identified was to increase seal temperature—and this was driving the over-sealing condition we were seeing.
 
This combination of a test protocol focused on detecting weak seals and an operator mindset focused on creating strong seals was destined to produce an unsatisfactory customer experience.
Plant Floor Quality Management 
Backed by this data, we implemented stricter controls on seal temperature adjustments and implemented a training campaign focused on educating operators and mechanics on the customer impacts of seal quality and seal temperature.
 

The Power of ProFicient

There’s real power in this feature of the software. By channeling operator responses into a pre-set list of options, you can start to get good trend data. By structuring the assignable causes and corrective actions appropriately, you get a few categories with multiple responses—rather than getting 40 categories, each with one or two occurrences. 
 
Often, when you’re looking at 40 categories, you’re not actually getting 40 different root causes; what you’re getting is more akin to having 40 eyewitness accounts of a single incident. Harnessing the ProFicient software, you can convert that scatter-shot data into focused data that have value and purpose.
 
Finally, after all these tests, we went back to our Assignable Cause and Corrective Action code data, again leveraging a Pareto chart, and were able to quantify the improvements we had implemented.
Pareto Chart
Using the built-in data collection and analysis tools of ProFicient quality management software, we turned the qualitative into the quantitative and isolated the root cause of our problems—improving our manufacturing processes along the way and ensuring consistent, high-quality products for our customers.
 
For the purposes of this blog, we did not go into details regarding the software. To see step-by-step details of the ProFicient software in action, please check out the Tales from the Trenches video series here.
 
Feel free to check out the other blogs in this series:
Take advantage of the technology at your fingertips today: contact one of our account managers (1.800.772.7978 or via our website) for more information.
 
 
Ian Farrell
By Ian Farrell
Lead Consultant, Key Performance Quality Consulting, LLC
See Full Bio


 

SUBSCRIBE TO OUR BLOG

Never miss a post. Sign up to receive a weekly roundup of the latest Quality Check blogs.

Archives