Nowadays, the theory of reliability is widely used in the construction industry, and it is also used in the field of geodesy. By abstracting its position, it can be successfully transferred to systems that, it would seem, are not in a dynamic state. Let's take, for example, the points of the polygonal (leveling) network in the city. It would seem that such a network is in a static state, but over time it undergoes changes, that is, it is in imperceptible dynamics and its reliability gradually decreases.
 Reliability in the broadest sense of the word means the ability of a technical device (system, network) to operate without interruption (failure) for a given period of time under certain conditions. Such a period of time is usually determined by the time of execution of some task, which is carried out by the device or system and is part of the general operational task.
 Currently, the problem of reliability is becoming one of the central problems of engineering and management organization. Ensuring the reliable operation of all system elements is of primary importance.
 The modern interpretation of the term "reliability" in geodesy is associated with Baard's reliability theory, according to which reliability is defined as "the ability of the network to self-monitor against gross errors [1]. One of the indicators of reliability is the minimum amount of gross error that can be detected after statistical analysis of the corrections obtained after equalization. On this path, many difficulties arise, most of which remain unsolved, namely the impossibility of localizing a gross error, quantitative assessment of reliability, methods of designing reliable geodetic networks. The unsatisfactory state of reliability theory limits its practical application, but at the same time, the problem of product reliability is the main one in modern production and its importance is growing all the time. In this regard, there is an acute problem in highlighting new views on solutions to this problem in the field of designing and creating geodetic networks.
 A new scheme for constructing the theory of reliability of geodetic measurements is proposed. Its basis is a probabilistic determination of reliability, a system of tolerances, and methods of localizing gross errors. Reliability is defined as a probabilistic measure of the homogeneity of measurements that have passed control. Quantitative reliability is estimated by three indicators: differential, integral and the minimum gross error. The Neumann-Pearson criterion is recommended for establishing tolerance. This theory makes it possible to design geodetic networks with a given reliability.
Read full abstract