The use of a probabilistic approach is proposed in the modeling task. Several repeatable groups of the same type of operations are considered - the search for the cause of failure and its elimination. Based on the experience of previous repairs of the same type of car components, the distribution series of values Pm (probability of repair of certain nodes) and Pn (probability of failure of certain nodes) are set. The model takes into account the qualifications of engineers and technicians of the service station, the availability and capabilities of mobile and stationary monitoring and repair facilities. After determining the probability of correct troubleshooting and planning a certain algorithm of maintenance operations, the repair efficiency is estimated as the ratio of the probabilities of performing repairs for a given time and the average repair time. The desired probability density, given the multiphase nature of the process, is determined by the method of multiple convolution. The resulting model allows us to solve two reciprocal problems. Firstly, to determine the optimal repair time, provided that diagnostic and repair tools are equipped, as well as the qualifications of technical workers. Secondly, at a given repair time, determine the required characteristics of diagnostic equipment, maintenance facilities and personnel.