Some technologies (such as space exploration, air travel, and electric power generation) offer significant benefits to society, but have severe consequences of failure, as demonstrated by such accidents as with the space shuttle Challenger and the Chernobyl nuclear reactor. The need for highly reliable use of such technologies requires a disciplined approach to risk management in design, construction, and operations, particularly when several groups are involved in safety-related decisions. This interdisciplinary study is designed to incorporate organizational effects into classical reliability assessment. It will generate one quantitative measure of organizational effectiveness by assessing how such factors as hierarchical structure, incentives, and protocols for distributed decision-making can affect the physical risk. The national air traffic control system is the proposed application. Of particular interest is the formation of safety thresholds (the maximum allowable failure probability for all or part of the system) or the different groups involved; the way in which the standards evolve as system operation becomes routine (for example, drift toward laxity); and the effect of such variations on the overall system reliability.
