The purpose of the article is to investigate the reliability of an unmanned high-altitude module based on a mathematical model of the k-out-of-n:G system. An analytical model of the k-out-of-n:G system under two system failure scenarios is considered. In the first case, the system failure occurs after (n−k+1) elements failure. The second one examines the system failure depending on the location of the failed elements. The sensitivity analysis of system reliability characteristics to the shape of the lifetime distribution function of the components has been carried out. The impact of the coefficient of variation of the system elements lifetime on its operating probability without failure is investigated. Several machine learning methods are used to calculate reliability characteristics for arbitrary input data based on practically significant parameters. The accuracy of the trained models is expressed in terms of estimated mean values.