Davis Economics of Automatic Testing

1. Auflage 2013
ISBN: 978-1-62675-122-4
Verlag: BookBaby
Format: EPUB
Kopierschutz: Adobe DRM (»Systemvoraussetzungen)

E-Book, Englisch, 400 Seiten

Engineering is... 'the art of doing that well with one dollar-that any bungler can do with two-after a fashion'. This famous definition by Arthur Wellington*, the father of engineering economics, says it all. This book takes that premise, shows how it applies to testing decisions in electronics, shows why full life-cycle costs need to be determined, shows how to make all of the calculations and shows how to present a case to senior management in terms they understand. The book content is the 1994 second printed edition-it remains up to date because the economic concepts and methods described in the book have not changed. There are a few additional notes which have been added for clarity. *Arthur Wellington, The Economic Theory of the Location of Railways - Pub 1887

Autoren/Hrsg.

Davis, Brendan

Leseproben

1. Introduction and overview Notes in these boxes contain comments that were not part of the 1994 printed book. They have been added in places to add clarity or to simply comment on something in the main text.
It is widely recognized that automatic testing of electronic components and sub-assemblies has a number of distinct advantages over more manual test methods. Probably the biggest advantage is in the area of economics, since although automatic test equipment (ATE) has a relatively high capital cost its overall operating cost is usually significantly lower than the alternatives of manual testing or no testing. This alternative of ‘no testing’ is what most people would like to achieve. If we could guarantee that the design was error-free, that every component purchased would meet all of its specifications and be assembled and soldered correctly and that boards in a system were interconnected properly, we could eliminate testing. Assuming that the design required no adjustments or calibration, the product should be shipped to a customer directly off the production line or from the system integration area. In the real world, if we did this, our field-service costs would be astronomical—they are probably high enough even with all the testing we do anyway—and we would rapidly go out of business because of a terrible reputation for poor quality and reliability. Testing, then, can be thought of as a cost-avoidance activity. Within the production cycle of an electronic product there are a number of stages at which testing or inspection can be performed. It is generally accepted that the earlier in the process a fault is found, the less it costs. It should be clear, however, that if all the testing that could possibly be done is in fact performed, then probably too much would be done, resulting in a less than optimum operation. Economically speaking, more testing than is necessary is just as costly as not doing enough. Quite naturally there will be differences between available test equipment: not only the differences between the products of the various vendors but also differences within the product range of an individual vendor. Deciding which product is best suited to your requirements can be a very difficult task, especially when some of the more important differences, such as software support and the credibility of the vendor, are difficult to measure. The major purpose of this book is to… 1. Outline the advantages of automatic testing relative to the alternatives of manual testing or not testing. 2. Highlight the major cost and savings areas, showing how to calculate potential savings. 3. Highlight the important areas for consideration when evaluating automatic test equipment. 4. Discuss the trade-offs to be considered when trying to optimise the testing strategy for a product. 5. Show how to perform a financial analysis of an intended investment in ATE and present the results in a way that financial and upper management can understand.

Throughout the book the emphasis will be on economics rather than on the technical aspects of ATE. Inevitably, however, some reference needs to be made to technical matters since they affect the economic performance of a given type of system. In particular, the emphasis will be on minimising costs and maximising productivity rather than simply lowering costs and increasing productivity. Any ATE system, no matter how poor and no matter how poorly used, should perform more economically than the manual testing alternative, but is not necessarily the optimum solution. To optimise the testing operation, four major objectives need to be accomplished… 1. The development of testable designs. 2. Selection of the optimum testing strategy. 3. Selection of the most cost-effective equipment. 4. The effective management and control of the testing department.

The first three objectives are addressed by this book. The fourth objective is outside the scope of this book and probably deserves a book all to itself to do justice to the subject. Another objective to be kept in mind when considering the others is the maintaining, or better still the improvement, of quality and reliability of the product. The effects of the rapid growth in the use and complexity of semiconductor devices referred to in the Preface is having another effect. In some areas, products are becoming more and more alike. The chips are available to everyone and it is easy to copy an innovative design. The result is that companies have to search for other ways to differentiate their products from those of their competitors. An increasing form of differentiation is that of quality and of after-sales support such as field service. More and more, it will be the company with the better reputation for quality and service that will gain market share over its rivals. The testing strategies selected must therefore make every attempt to improve quality and reliability to make the field-service job easier to perform in an efficient manner. Another issue that has gained substantially in its importance over the past ten or twelve years is that of ‘time-to-market’. Getting new products to market on time, and hopefully ahead of the competition, has become increasingly important. This has been driven by the intensification of world-wide competition that has tended to shorten product lifetimes as companies try to leap-frog their competitors with better products at lower prices. Possibly one of the best examples lies in the personal computer industry. Technological changes have been rapid and price reductions have been dramatic. The ‘laptop’ computers that enjoyed premium prices just two years ago (1990) are now virtually obsolete as the smaller and lighter ‘notebook’ PCs have caught up and surpassed them in terms of performance. My own laptop, just two years old, is currently available for one third of its original (discounted) price. A colour notebook launched in March 1992 at a list price of £5,500 is already available for less than £3,000 and it is only July 1992 as I write this. Here is an area that is clearly out of date! Laptop computers, with the exception of a few very high-end units, are now a fraction of these prices and have many times the features and performance levels. This is a classic case of electronic products getting cheaper and more complex at the same time.

Note that the prices quoted are in pounds sterling! I recently (1993) read an article about Michael Dell, founder of the Dell Computer company. In the article it mentioned that their new product design cycle is already down to eight months and the oldest product in their catalogue is just fourteen months old. A number of direct selling PC manufacturers have recently discontinued all models using the 386 series of processors. They only make 486 based units with a starting price below £1,000 and the 586 chip is expected soon. By the time this book is available this will all be stale news. The prices I have quoted will seem very high and the 486 will possibly be on the way out. This pattern, to one degree or another, is common throughout the electronics industry—shorter product life cycles—increased use of technology—lower prices and better quality. This is all good news for the end user, but the cause of many headaches for the engineering, manufacturing and test engineering people. 1.1 What cost quality? More often than not quality, or more particularly high quality, conjures up thoughts of a Rolls-Royce, a Mercedes or a Cadillac. As a result, many people equate high quality with high cost. However, if you think of quality as ‘fitness for use’ or as ‘conformance to specifications’, it becomes clear that it is possible to have a high-quality model of a much lower priced vehicle, or indeed a poor-quality example of one of the more exotic machines mentioned above. This concept of ‘quality’, epitomised by promotional slogans such as “this is the Rolls-Royce* among dishwashers”, is really only part of the overall meaning of quality. This use of the word is really referring to the ‘grade’, ‘class’ or ‘degree of excellence’ of a product or service. More generically, this can be referred to as the ‘quality of design’ since the grade selected will be a conscious decision of the design of the product. Looked at from the point of view of ‘fitness for use’, the quality of design will be a function of how well the marketing people have interpreted the real needs of the market-place, the concept of their product offering and the quality of the specifications laid down for the product. * The Rolls-Royce Car Company object, quite rightly, to the use of their name to promote other products. I object to it also but for other reasons. The ‘Rolls-Royce’ definition of quality is totally incorrect. There is always confusion between high-grade or luxury products and the correct ‘conformance to requirements’ definition of quality. Chapter 2 explains this more completely. Once the product is being manufactured, the...