Process improvement by poka-yoke

Michael Fisher, Freelance journalist based in Bradford, UK.

Shigeo Shingo is credited with creating the concept of zero defects and the techniques of poka-yoke (Japanese for mistake-proofing). The approach seeks to remove the causes of defects, or, where this is impossible, to inspect each item simply and inexpensively to determine that it passes the quality threshold - with no defects.

Thus, a poka-yoke device is any mechanism that either prevents a mistake or defect occurring or makes any mistake or defect obvious at a glance. It is the concentration on removing the causes of defects that is important. The inspection process is a backstop.

Shingo did make a clear distinction between a mistake and a defect. Mistakes are inevitable, defects are entirely avoidable and result from allowing a mistake to reach the customer. The goal of poka-yoke is to engineer the process so that mistakes can be prevented or immediately detected and corrected.

"The causes of defects lie in worker errors, and defects are the results of neglecting those errors."

-Shingo, 1986

Poka-yoke methods and devices can be very simple: for example, the old-fashioned visual indicators on shelves/storage so that storekeepers could see replenishment levels (rather than have to wait to be reminded by a card index or computer system).

When visiting the Yamada plant in 1961, Shingo was told of a problem that the factory had with one of its products, which included a switch with two push-buttons supported by two springs. Occasionally, a spring would get missed and the error might not be discovered until the unit reached a customer. An engineer would then be dispatched to disassemble the switch, insert the missing spring, and re-assemble the switch. This problem was both costly and embarrassing. Management would warn employees to pay more attention, but the missing spring problem would eventually re-appear.
Shingo suggested a solution that became the first poka-yoke device (Shingo, 1987).

In the old method, a worker began by taking two springs out of a large parts box and then assembled a switch. In the new approach, a small dish is placed in front of the parts box and the worker's first task is to take two springs out of the box and place them on the dish. Then the worker assembles the switch. If any spring remains on the dish, then the worker knows that he or she has forgotten to insert it.

The new procedure completely eliminated the problem of the missing springs. This satisfies both elements of the poka-yoke philosophy: it helps remove the cause of defects - and it provides a simple, visual "inspection" process. Many jigs and fixtures now incorporate poka-yoke elements. As with all good detection/inspection regimes, it is first, a very simple means of eliminating worker oversights and errors, and second, it ensures that inspection takes place at a point in the cycle when the cost of recovering from an error is low.

Shingo himself categorised three different types of inspection:

• judgement inspection;
• informative inspection; and
• source inspection.

Sometimes referred to as "inspecting in quality", judgement inspection is the traditional inspection process of identifying products that are defective before they are released for distribution. Modern quality thinking suggests that this is an inappropriate and expensive way of maintaining acceptable product quality.

Informative inspection uses data gained from inspection to modify the production process and prevent defects. Traditional statistical process control (SPC) is such a form of inspection. The kinds of checks built into zero-defect approaches can also often be informative inspection, since they are used subsequently to modify the manufacturing process.

Shingo built successive checks - where each operation inspects the quality of the previous operation - into his methodology, fearful of leaving inspection until it resulted in excessive scrap and re-working. Successive checks also allow the precise points at which errors occur to be identified - this allows faster and easier modifications to manufacturing processes and working methods. Each area thus becomes both a production stage and an inspection stage. Correctly handled, this can also increase responsibility and commitment of the workforce and can reduce the actual cost of inspection. As against typical sample inspection processes, it also means that every item is inspected (several times) and all defects should be identified.

An alternative is to allow self-checking of work by the operators who complete it. This provides even faster feedback! Both self-checking and next station checking are "after the event" checks - they determine that something that has been produced is defective. This means the product has to be scrapped or reworked.

Source inspection determines "before the event" whether the conditions necessary for quality production exist. Shingo comments: "It had dawned on me that the occurrence of a defect was the result of some condition or action, and that it would be possible to eliminate defects entirely by pursuing the cause" (Shingo, 1986). He then identifies that checking operating and working conditions before an operation commences may help eliminate defects.
Poka-yoke devices use source inspection to ensure that proper operating conditions exist prior to production. In some cases, devices are designed to prevent production from occurring until the necessary conditions are satisfied. This is a common process within automated manufacturing systems where a manufacturing cycle cannot start until all components are detected as being in place.

Source inspection, self-checks, and successive checks are used together to gain maximum feedback in the shortest time so that the manufacturing process is both understood and managed. Although the immediate result is that defects are identified and prevented from progressing, the real aim is to design out future defects through improved processes. Source checking is preferable because it eliminates defects but is not always possible. Then, self-checking is preferable since it provides information faster - and enables more accurate identification of defects. Shingo refers to his control process as "zero quality control". The building in of these inspection mechanisms should result in the inspection process being an integral part of production. Thus, it is not an "overhead" on production and has zero cost. Shingo decried SPC as providing only partial (sample) inspection and for being an "afterthought"; merely identifying defects and allowing for the same number to occur in future runs without any element of built in prevention. This is probably a little harsh, since SPC is often used as the basis of process analysis and review: the prevention may simply be one more step removed, and less automatic in its inclusion.

It is reasonable to suggest that since SPC usually involves significant intervals between the sampling of batches, feedback will be faster with source inspection and informative inspection than under traditional SPC-based inspections. However, this does not necessarily mean that Shingo's methods are systematically faster than SPC at ensuring corrective actions are undertaken. The willingness and ability to take corrective action is a function of the attitude and commitment of both managers and workers, not an intrinsic attribute of a particular approach to inspection or quality management.
Shingo advocated the practical application of zero defects by good engineering and process investigation, rather than the slogans and exhortations that have been associated with the quality campaigns of many Western companies. He argued that displaying defects' statistics was misguiding and demoralising. Instead, the positive results of improvement should be announced and displayed. Culture thus follows the engineering and is a by-product of it, rather than cultural change being the process by which improvements are realised.

The much vaunted Japanese quality miracle arose from the application of many thousands of simple poka-yoke devices. Each one is relatively simple - but together their effect is enormous.

This is a shortened version of an article originally published in Work Study; Volume 48, Number 7, 1999.

The author was Michael Fisher.

Copyright © MCB University Press.

You may also find the following Emerald Fulltext articles of interest:

• Control triggers: a control concept come of age by Irvin T Nelson and Richard L Ratliff describes control triggers as signals which initiate the right activity to occur at the right time in a process, and discusses them as a previously unrecognized category of internal control methods.
• An inspection and repair strategy in automated assembly systems by H Yamashina and H Mizuyama shows that the number of short stops in assembly lines can be reduced by installing inspection stations and repair stations strategically.
• Order completion accuracy - a pragmatic approach by Michael Galley extends the view that order picking is just one of several activities undertaken in delivering goods to customers, and all activities need to be examined to give a true picture of error rates in meeting customers' orders.