In order to organize transport systems work more efficiently it is important to analyze and to optimize them on the base of corresponding mathematical models. Often necessary data gathering is complicated with different factors influence, what leads to small data sizes and wrong interpretation using classical methods of analysis. In such situations it is convenient to use resampling method of statistics that gives robust estimators of parameter of interest, taking as efficiency criterion the mean square error. In the present paper the resampling-approach implementation is considered to a task of storage control theory in transport systems. Suppose we have two simple independent renewal processes: demand {Xi, i=1,2,...} and isupply {Yi, i=1,2,...}, where {Xi} and {Yi} are the sequences of unnegative independent random variables, each sequence with its own common distribution. The initial inventory level K is known. Our purpose is to estimate the probability of the shortage absence, it is the probability, that the m-th demand comes later, than the m-K -th supply.