ABSTRACT
As a generality, technical people want to test to obtain knowledge, whereas commercial people will test only when there is some pressure to do so, In an age of cost-cutting and streamlining of production, it may seem that testing is an unnecessary expense, but the reverse is true, since alongside an awareness of cost has grown an increasing customer awareness of quality, The consequences that arise when testing is omitted are illustrated by the following examples,
Some years ago a colleague who served on several SI committees once attempted to buy a radio while he was abroad, The young assistant took one off the shelf, switched it on, and it didn't work, She then unpacked one from its box just as it had arrived from the factory, and that one didn't work either, Eventually she found one that did and was surprised when he declined to buy iL What astounded him was that these products could be made and packed without any testing whatever to prove fitness for purpose until they reached the point of sale,
However, testing the product may not be sufficient, since in a complex product such as a radio, the reliability of the assembly depends on the reliability of each of the individual components, This was brought horne to a supplier who asked a buyer what failure rate he would accept, The curt answer "zero" failed to convince him and 0,1% was suggested, "No, zero," was the reply, "But that implies testing every component," said the supplier, "Exactly," answered the buyer, In practice, of course, virtually zero reject levels can be achieved without 100% testing by extremely tight control of the process, but the story illustrates the point, The increasing demand for product quality brings in its train a requirement for component reliability, and that implies component testing also,
Staying with this frictional analogy, the inclined plane method of measuring friction would give an apparent coefficient of friction, since conditions are not tightly controlled (velocity, for example, cannot be specified) and there is little chance of relating the result to other conditions.
