2 

 = 1 



It is assumed in this study that a-] = 1 ; that is, that the heavy-lift 

 system is always in its fully usable state at the beginning of a mission. 

 Considering the system will be used for construction, this is a reasonable 

 assumption. As a consequence of this assumption, it can be seen that the 

 heavy-lift system of this study does not possess the operational characteristics 

 of static alert systems; that is, those systems which are kept in readiness for 

 emergency use at some future time in a role of search, defense, orretaliation. 



Since the availability vector reduces to a-| = 1 , the effectiveness 

 equation reduces to 



2 

 J = 1 



dj Cjk; 



Dependability Matrix. The above assumption on the availability 

 vector (i.e., a-] = 1 and a^, = for n > 1 ) reduces the dependability matrix 

 to a Ixn row vector. 



It is recognized in this report that great and unavoidable uncertainties 

 are part of evaluating system dependability. Unfortunately, it is not possible 

 to confront these uncertainties with traditional statistical techniques. For 

 instance, the determination of confidence limits requires the examination of 

 samples drawn from a well-defined population subjected to a well-understood 

 environment. In the case of the heavy-lift system, there is no sizeable sample 

 to examine for calculating the usual measures of reliability. 



Since there is no possibility of accurately determining the dependabil- 

 ity (failure rates, etc.) of the candidate systems, a less formal technique is 

 necessarily used. Simplification of the problem is possible by assuming there 

 are two states of interest after the system has begun a mission: (1) the fully 

 operable state and (2) the fully inoperative state. These two stochastic 

 system states simplify the effectiveness vector from 



36 



