Statistical Process Control – Not just an investment in your business – An investment in your Employees, and your Corporate Culture!


How can a robust SPC program affect employee development and cultural improvement breakthroughs when implemented across all lines on the factory floor?

The utility and business value of using on-line, real-time Statistical Process Control to improve and understand process performance, reduce waste, and maximize quality, is extremely well documented. Walter Shewhart of Bell Laboratories developed control charts in the 1920’s, so this set of tools is hardly a new addition to the quality toolbox.

While the technical logic of using Control Charts specifically, and SPC in general, may be familiar to you, this blog entry focuses on the employee development and cultural improvement breakthroughs that are accelerated when a robust SPC program is implemented across all lines on the factory floor.

Data Utilization

Data is almost always collected, during production, to ensure that that a given batch is acceptable for shipment. Employees spend a lot of their time collecting, documenting, and submitting data for review. While employees probably understand that this data is allowing batch release, they must often wonder… is anything being done with this data, to help make my line run more smoothly? Is anyone noticing how many adjustments I’m making, or how close I’m running to my specification limits – or perhaps, is anyone noticing how many times I might be exceeding specification limits (usually with overfills that are automatically “approved”)? And did anyone notice how much better (or worse) this last batch ran compared to the prior ones?

With statistical process control software, production performance history can be easily reviewed for long-term trends. Technical or quality staff can be alerted to unusual process performance well ahead of any specifications being exceeded. And by noticing such issues immediately, closer to the event, the ability to talk to line employees or review batch records for any unusual events is MUCH easier, compared to waiting for a customer complaint or some other OOS result to occur before seeking to understand what is being seen on the “shop floor”.

It must be disheartening to an employee to spend all this time collecting data, only to have their data go into a deep dark file, with no obvious actions taken from it. Conversely, to see their data being used to justify improvements or perform improvement investigations energizes employees, as they now believe that this effort and data collection is leading to real action/support by their Management.

Employee Awareness and Ownership

Any enlightened Manager believes that the vast majority of employees want their business to succeed, and want to “do the right thing”. One sure way to encourage and promote this behavior and attitude is to give employees an easy way to understand the link between their everyday work, and the performance of the business.

With statistical process control software, it is extraordinarily easy to quantify the business benefit ($ saved) of “running to target”, vs. allowing a process to drift, or to let an out of control condition exist without taking steps to correct the issue. In real time, a line technician can receive information like how much money was saved TODAY by running 1 more gram closer to target than occurred just yesterday. Or, they can see the total theoretical out of specification units, and see if they are trending in the right direction and thus be informed about the likely improvement in the rate of customer complaints, or improved consumer satisfaction, due to their ability to reduce the number of OOS units.

All of these results can easily be summarized across lines in the factory and link to corporate savings targets, ppm defect reduction targets, etc. for which a business has set goals for this fiscal year. Employees who go home EVERY DAY, knowing how their performance impacted business results, is more connected to their business’ success and is a stronger owner in the quality of their work. Armed with such information, they are empowered to approach their managers with improvement ideas, or to ask questions as they seek to understand how they can do better – and how inputs which might contribute to operational problems are being dealt with in a proactive way.

Respect for Specifications

Perhaps the most poorly appreciated positive of the real-time and wide-spread use of statistical process control is the respect that this tool brings for the development of truly RELEVANT specifications, which significantly improves all employees’ respect for those specifications.

If a process exhibits an inability to be “capable” of routinely meeting product specifications – whether it’s product weight, cap torque, label skew, whatever – a critical first question that must be asked (but rarely is asked, without the use of SPC) is: “what is the basis of these specifications”? For instance, is there a real customer need for all units to be within these specifications? Or is this a “rule of thumb” specification range that became cemented over time, and does not link in reality to the product and process being run?

All too often, specifications are set without consideration for what is truly noticeable to customers and consumers. Even more often, they are set without any regard for what an equipment train is capable of delivering – when you add the variation that naturally occurs in materials, process variability, test method variability, etc. Therefore, the risk of running outside of these “arbitrary” specifications is significant. Then, when a specification is exceeded, sometimes routinely, it becomes “automatic” to gain approval for a deviation to the specification, that allows a line to continue to operate, and reduce downtime. What kind of message does this send to line employee? Some specifications are “ok” to exceed, while others are not? Which is which? And how is anyone held accountable for a requirement to meet specifications in this environment?

It’s easy to picture employees going home every day shaking their heads in confusion when they are told that “specifications” are critical to remain within, yet excursions outside of certain specifications are routinely accepted. Instead, we would be supporting a high compliance environment if data was used to examine and challenge the justification for specification ranges – and investments were made to ensure that these “valid” specifications are indeed routinely met. Occasional deviations will still happen – but will be much more the exception rather than the rule of daily life!

Employee Recognition for Improvement

Enlightened Management wants to recognize their employees positively when they make positive contributions – or alternatively, give employees important improvement feedback if expectations are not met. Without data, how does one know which is which? When is an issue truly an unexpected issue, vs. natural variation? How do you know the difference?

Below is an example of an attribute (p) chart which shows % defective units for the manufacture of a product, and charts the % defect rate (on they axis) across several sample subgroups over time (on the x-axis ). For the purpose of this discussion, let’s assume that each subgroup corresponds to samples taken during a shift of production.

Applying the rules of control charting, none of this data for any of the “shifts” of production are outside the control limits. However, without a control chart mentality, it would be easy for uninformed management to draw some very incorrect conclusions, like: “why did the teams on shift 1 and 5 run so POORLY – what’s WRONG with them?”, or “hooray for shifts 6 and 10, let’s have a celebration lunch for their GREAT performance!”. When in reality, there is no statistical difference in the performance of ANY of these shifts! The teams on shift 1 and 5 will feel beat up – inappropriately – and shifts 6 and 10 feel great NOW, but what about when natural variation comes in and have their “in control” performance of 20% defects? What then? And what if the same people ran on shift 1 and 10? How confused must they feel!

All of these impactors on a company’s culture can be enhanced through the proper training, use, and communication of statistical process control techniques.

Has your organization experienced cultural benefits through the use of real-time data creation and management? If so, feel free to share your success stories with us!


Lauren Wilson


Comments