Aim. Calculations are an integral part of the development of any complex technical object. Normally, they are subdivided into the calculations to confirm product operability (kinematic, electrical, thermal, strength, hydraulic and pneumatic systems analysis, etc.) and calculations to confirm its dependability (calculation of reliability, longevity, maintainability, storability and other indicators). As it is understood and provided in statutory documents, dependability calculation involves procedures of identification of an object’s dependability indicators using methods based on their calculation using reference information on the object’s components dependability, on the dependability of analog objects, on the properties of the materials and other information available at the time of calculation. However, in the case of development of unique highly vital systems, obtaining statistical data for dependability calculation is impossible due to two conflicting conditions, i.e. the limited number of produced objects and the requirement of high accuracy of the input information. Nevertheless, in the author’s opinion dependability calculations must be performed. The only question is how to calculate the dependability and what such calculation should mean.Methods. In the classic dependability theory, the conventional understanding of probability of no-failure is the frequency of failures in time, yet for unique highly vital systems the failure rate must tend to zero over the entire period of operation (preferably, there should be no failures at all). For this reason the concept of “failure” in the context of unique highly vital systems should probably be interpreted not as an event, i.e. any fact, which as a result of experience can occur or not occur, but as possible risk, i.e. an undesirable situation or circumstance that is characterized by the probability of occurrence and potentially negative consequences. Then, an event in the form of a real or potential failure in operation can be associated with a risk in the form of probability of failure with negative consequences, which in terms of the consequences is equally unacceptable with regard to unique highly vital systems. In this case dependability calculation can be reasonably substituted with risk assessment, a process that encompasses risk identification, risk analysis and comparative risk assessment. Thus, risk assessment enables the achievement of the target dependability directly by substantiating the stability of manifestation of a specific product’s properties and not indirectly through undependability caused by failures of analog products.Results. The paper shows the procedure of risk assessment for unique highly vital systems. Using the example of a mechanical system with actuated parts represented by a spacecraft single-section pivoted rod the risk assessment procedures are shown. The feasibility of risk assessment with the use of design engineering analysis of dependability is demonstrated.Conclusions. It is shown that the absence of statistical data on the dependability of analogs of unique highly vital systems does not prevent dependability calculation in the form of risk assessment. Moreover, the results of such calculations can be a source and guidelines for adopting design and process engineering solutions in the development of products with target dependability indicators. However, legalizing the method of such calculations requires the modifications of the technical rules and regulations to allow for dependability calculation by other means than with the use of statistical data on the failures of analogs.
Disclosure risk assessment in statistical data protection
