The application of logistic nets is an integral part of large-scale tasks management. The nets are designed to be maximally stable; however, the risk of link damage between the elements of the net cannot be excluded. The possibility of reconstructing the damaged link on time, thus minimising the losses, is a task of high priority. This paper discusses using first-order rules for solving that task. To accomplish searching for first-order rules, the FOIL (First-Order Inductive Learner) algorithm is studied and employed. As a result, a set of first-order rules is obtained that partly describes the net used, based on the information about the properties of the logistic net. In the course of the experiment, damages of one or several links in the logistic network were created artificially. A model of the net constructed with the FOIL algorithm, was successfully applied to searching for alternative paths.