Demystifying Process Capability

Process capability is often viewed as a maze of perplexing concepts, shrouded in technical terminology only comprehensible to the initiated.

Picture this scene: A visibly distressed boss storms into the quality director’s office, disturbed by a crucial issue. Their main client’s assembly line has come to a standstill because the last batch of XYZ-50s delivered didn’t fit into their assembly fixtures. The quality director points out that the client requires a Process Capability Index (C_pk) of 1.33 or higher. However, the computation is based on an assumption of normal distribution, which doesn’t hold for the XYZ-50 as it has a skewed distribution. This technical conversation might leave others baffled.

The truth is, process capability, at its core, revolves around quality. It answers a straightforward question: “Can you fulfill my requirements?” Ideally, the client would appreciate a simple “yes” or “no.” However, this can be challenging due to factors such as imperfect inspection, inherent process variation, measurement inaccuracies, and misunderstanding of requirements.

Despite these complexities, the answer to the client’s question is rather straightforward: no. Due to the aforementioned reasons and more, guaranteeing a product or service that always meets the client’s requirements is an impossibility.

What then? The optimal strategy might be the most radical one: Honesty. Inform clients about the number of items that might fail to meet the requirements. This direct approach gives clients the information they seek. It works for both variable and attribute data. If control charts are being used, this estimate can be derived from the process average or from the process average and standard deviation.

In case of a highly efficient process, we can assure the client of near-perfect quality. One way to articulate this is through parts-per-million quality statements. For instance, “Our return rate on this item is three returns per million items in service per year.” This simple statement can easily be understood by most people.

If the process is less capable, providing the anticipated number of defective items could come as a surprise to both the workforce and the client. This revelation might be the push needed for quality improvement.

For high-volume production, expressing process capability as expected defects can be enlightening. While a defect rate of 1/10 percent might sound impressive, for a can line that produces over 1,000 cans per minute, this translates to 1,440 defective cans per day. In such cases, parts per billion quality might be required.

If a process isn’t statistically under control for unforeseen reasons, it’s impossible to precisely state the process capability. The optimal option would be to inform the client about the expected defects based on historical data.

Clear communication is key to excellent client relations. The most straightforward way to convey the message is by informing the client about the anticipated quality of product or service, in simple language.