Systems Reliability and Failure Prevention

Common sense (reinforced from time to time by advertising) tells us that the reliability of our vehicles, appliances, and services (utilities, banking) depends on the reliability efforts made by the vendor. These efforts, in turn, are largely dependent on what customers demand. Our toleration of unreliable equipment and services has radically diminished in recent years and, in response, most vendors have been able to improve reliability. In this chapter, we examine the tension between economy of resources and the (sometimes) very high cost of failures. We look at applications associated with high reliability requirements or expectations and analyze failures experienced there. We concentrate on the management and organizational aspects of the failures; physical causes of failures will be discussed in later chapters.
We value safety and reliability, and we take pride in our accomplishments in these areas. When we see deficiencies, we demand that those responsible take corrective action. Frequently these demands are pressed by lawyers (who do not seem to mind that there are occasional lapses of responsibility at high corporate levels). Our professional schools teach good design practices, government at all levels enforces safety requirements in our buildings and vehicles, and developers of equipment for critical applications have standard practices in place to avoid known causes of failure. One way of evaluating the results of these efforts is through accident statistics: for example, the death rate from nonvehicular accidents declined from 94 per 100,000 population in 1907 to 19 in 1997. Vehicular accidental...