## 71 Development of Reliability Engineering

Some of the earliest developments in reliability engineering occurred during the Second World War. The Germans had problems with the reliability of the V-1 missile (Bazovsky, 1961). The project team leader, Lusser, has described how the first approach taken to the problem was based on the argument that a chain is no stronger than its weakest link. This concentrated attention on the small number of low reliability components. But this approach was not successful. It was then pointed out by a mathematician, Pieruschka, that the probability of success p in a system in which all the components must work if the system is to work is the product of the individual probabilities of success p,:

This drew attention to the need to improve the reliability of the many medium reliability components. This approach was much more successful in improving missile reliability. Equation 7.1.1 is known as Lusser's product law of reliabilities.

On the other side of the Channel, Blackett (1962) was drawing attention to the significance of Equation 7.1.1 to military operations in general:

In the simplest case of air attack on a ship, the four main probabilities are (1) the chance of a sighting, (2) the chance the aircraft gets in an attack, (3) the chance of a hit on a ship, and (4) the chance that the hit causes the ship to sink.

This work was the beginning of operational research.

The US armed forces also had serious reliability problems, particularly with vacuum tubes used in electronic equipment. Studies of electronic equipment reliability at the end of the War showed some startling situations (Shooman, 1968a). In the Navy the number of vacuum tubes in a destroyer had risen from 60 in 1937 to 3200 in 1952. A study conducted during manoeuvres revealed that equipment was operational only 30% of the time. An Army study showed that equipment was broken down between two-thirds and three-quarters of the time. The Air Force found that over a 5-year period maintenance and repair costs of equipment exceeded the initial cost by a factor of 10. It was also discovered that for every vacuum tube in use there was one held as spare and seven in transit, and that one electronics technician was needed for every 250 vacuum tubes. These studies illustrate well the typical problems in reliability engineering, which is concerned not only with reliability but also with availability, maintenance and so on.

From these early beginnings the study of reliability has become a fully developed discipline. It has received particular impetus from the reliability requirements in the fields of defence and aerospace, and electronics and computers. Perhaps the most spectacular example is the moonshots, which depended crucially on reliability technology.

One of the main fields of application of reliability engineering has been in electronic equipment. Such equipment typically has a large number of components. Initially the reliability of electronic equipment was much less than that of mechanical equipment. But the application of reliability engineering to electronic equipment has now made it generally as reliable.

Another area in which reliability engineering has been widely use is nuclear energy. Methods have had to be developed to assess the hazards of nuclear reactors and to design instrument trip systems to shut them down safely.

In the UK, work on reliability of nuclear reactors has been done by the UK Atomic Energy Authority (UKAEA), originally through its Health and Safety Branch and subsequently through the Safety and Reliability Directorate (SRD), which runs the National Centre for Systems Reliability (NCSR) and the Systems Reliability Service (SRS). The latter operates a consultancy service on industrial reliability problems and a failure data bank.

The development of reliability engineering and of loss prevention in the chemical and petroleum industries has been described in Chapter 1.