1.1. Scope#

The reliability.space handbook was developed during an ESA study aiming at the development of a new reliability prediction methodology (NRPM) for space applications. The objective of the handbook was to provide a reliability methodology that can be used for space applications, without the limitations and shortcomings of existing reliability prediction methodologies that are mainly focussed on ground applications.

The term reliability prediction (RP) is here to be understood as the process, or outcome, of predicting the reliability of a system or its components, i.e., the probability of success. In the more general context of RAMS analyses, making use of qualitative and quantitative techniques, RP is restricted to providing quantitative estimates for the probability of success or failure. The focus of the methodology presented herein is on reliability in this narrow sense. Specific requirements of availability, maintainability or safety analyses have not been considered during the development of the methodology and are out of the scope of this handbook, even though the probabilistic methods provided may of course also be used to support these analyses.

For the purpose of the handbook, the term space applications refers to RPs for systems that operate in space, e.g., a spacecraft. While parts of the methodology may also be used to model end-to-end systems including the ground segment, the handbook does not cover aspects specific to ground applications, e.g., related to the effect of ground environmental conditions. In terms of technology coverage, the methodology is presently limited to unmanned spacecraft technology.

Even though the handbook targets decision support, the decision-making as such is left to the user; reliability predictions provide support to the decision-making, for example in a trade-off exercise. Also, the practical implementation and development or choice of suitable software tools is out of scope for the handbook. As a general rule, the models provided allow the development of tools without any restrictions regarding the information needed to implement the models.