Inspection and the Operation

 

There is a far more subtle factor to be considered when inspection is used in the way described above, and that is the effect on the thinking of the operators. Whether we like it or not, it says, "There is no need to strive to do the job properly—we have arranged for someone to pick up all the mistakes that you make. This was brought home strongly to me as a. Young newly appointed manufacturing manager. There were quality problems in my plant which were causing a high rate of field failures, and I did not know how to tackle the situation. Inspection had been beefed up by my predecessor to the point where there was an inspector to every five operators, but the field-failure rate of the product did not drop. Walking down the production lines one day I saw a component lying on the floor. I picked it up and gave it to the operator from whose bench it had fallen with the comment. “You will need this.” “Oh, no”said the operator, "It has nothing to do with me— it's his responsibility," pointing to the inspector sitting immediately behind him. As a result of this experience I reorganized the shop floor and removed all inspection from the production lines. I also changed the role of quality control from inspection to that of process monitoring. The responsibility for quality was placed squarely back on the operators. With some other minor actions, field failures fell in a few weeks to one tenth the original figure. We did not have a quality problem we had a management and organizational problem.

What then is the task of inspection in the well-controlled process? Certainly not to decide whether the product is good enough to ship or not, for that will be taken care of through excellent process control. Instant, inspection is needed to develop data. These data are used first to assure that the process is indeed within the control limits, and second to initiate any preventive actions that may be necessary to maintain or improve the process control.

For example, suppose that the product has a highly polished finish. In an uncontrolled process the inspector would check every item for defects in (ho polish, returning those that did not pass his inspection criteria for re-finishing There may well be some limit set on the number of items that can he returned without some investigation as to why these defects are being generated, but there is usually no attempt to eliminate the problem There is "an acceptable defect level." In this case the accuracy of the inspection becomes very important. If the limit is set at 1% but the inspector is only 70% efficient, there may well be a number of products shipped that have scratches or paint dribbles The inspector is the final arbiter, he is solely responsible for the quality of the finish and will be blamed for any defects. Yet it is impossible for him to ho 100% efficient, and from the very start we know that he will miss some errors.

 

 

Rework

 

What can be done to make the product quality right when it is found to be faulty? The product can certainly be reworked, but this is expensive and in many cases will result in nothing more than a cosmetic fix. For example, the cracked casing on a lawn mower that was dropped while it was being as-assembled may be found at inspection and replaced, but its bent shaft or out-of-line bearings are unlikely to be discovered and may ultimately result in a short life for the machine. In other words, rework will hide from the customer the fact that a problem existed in the product quality, because the cosmetic defects will be fixed. It will also hide the fact that problems exist in the manufacturing operation—in this particular case, the way the lawn mowers are handled and stored—which have to be reviewed and improved. Rework removes the pressure on production to fix the problems promptly because the products are eventually being shipped, but in many cases this means that the problems are never fixed. After all, if we believe that inspection can find every defect in every product and there is a group of people whose job is solely to fix products that fail inspection, why bother to prevent the problems from occurring? In many cases the rework department is nothing more than a catch-all that hides from management the fact that quality problems exist and removes any pressure to eliminate those problems.

 

 

Rework and Reliability

 

It would be naive to say that no rework should ever be carried out, although some companies are in fact making this part of their overall objective. How­ever, rework should never be carried out without a thorough understanding of what is to be done. The rework methods should be clearly defined, the operators well trained. Above all other considerations, it must be determined that the rework will not impair the life, or performance of the product. Finally rework must never be carried out lightly; me must recognize that it indicates a bacillus somewhere in the process and the basic cause of that failure must be found and eliminated.

Any rework, no matter how small or simple, is an unnecessary expense that demands immediate corrective action.

 

 

What is the Objective?

 

Our aim must be to make the highest quality product that we posibly can. It must perform perfectly, be completely reliable, have a superb appearance, and last forever. It is unlikely that we will ever achieve this aim, but if our objective is lower than this it is unlikely that we will ever achieve more than mediocrity Perfection must always be the ultimate quality requirement and our overall objective.

Living in a practical world we have to consider the many factors that will determine the lowest level of product quality that can be shipped, that is, the quality level that is acceptable to the customer and produces a saleable product. This is a much more difficult level to define and requires that many facets of our business be examined closely.

• At [he least, the product must perform properly according to its advertised claims.

• The sales price must be competitive.

• The reliability must be at least as good as the competing products.

• Its appearance must attract the customer.

• It must be convenient and safe to use.

 

 

The Specification Level

 

The specification level defines the absolute minimum level of acceptable quality and nothing less than this quality must ever be allowed out of our plant. We have to set our goals much higher than this or we will never get away from constant quality problems and bickering over specification requirements.

Setting such a goal is like attempting to fly over a 20,000-fooi mountain 30,000 feet; these minor variations will be a nuisance but will not affect the outcome of the flight. There is plenty of tolerance for unplanned and unex­pected variables, and there will always be variable forces acting on the aircraft just as there will always he variable items in our manufacturing process Although we must minimize these variables, they will always exist to some degree and there must be sufficient tolerance in the process to accept them without causing the product quality to deteriorate below the specified level.

There are many product specifications already in existence, ranging from those applicable to military and aerospace hardware to those developed by the various standards organizations and industry organizations. It is rarely neces­sary to start from scratch and write a new one. Yet most suffer from the same problems.

 

• They attempt to specify the barest minimum that can he accepted on the incorrect assumption that good quality will make manufacturing more expensive.

 

• They do not define what should be done when the product fails to comply with the specification.

 

. The author frequently sits on committees that write product specifications, and invariably a majority of its members have one idea in mind and that is to dilute the specification to the absolute minimum quality level in an attempt to reduce the cost of manufacture. Typical is the case of the criteria for acceptable solder joints in electronics. We would all like to see that every joint is perfectly made with a complete soldered connection around 360 degrees of the lead, which would guarantee a sound reliable joint. Yet, at every meeting on the subject, the argument is put forward that 180 degrees of joining is "all right " The argument then goes on to say that if a specification calls for 360 degrees of solder around the lead, then all of the joints having a lesser amount of connecting solder will have to be reworked to make them comply with the requirement and this will push up the labor costs. Here is a case where the objective is to do the least that is necessary to gel the product out, not to provide the ultimate quality and reliability. The simple fact is that it costs no more to make a good joint with a 360 degree fillet than it does to make one with a 180 degree fillet.

 

Quality and the Future

 

Because it is unlikely that we can ever know what our competitors are planning to do, our struggle for quality can never end. There will never be a clear objective: what will be adequate today will not be competitive tomorrow. Not too many years ago we were quite happy if the tires on our car lasted for 20,000 miles. We would never buy tires from a manufacturer today who could not guarantee a much greater life. Thus, we must recognize that quality is an open-ended objective; as soon as we close in on our target it moves farther away. We must also accept that our striving for quality can never end. Indeed, this constant struggle to improve, to compete successfully, is at the very heart of our society.

We can then add the following reference to the future, to our definition of quality.

Quality is a measure of the performance, reliability, and cost of our product. Our quality is adequate when our customers continue to buy our products.