Achieving sustainable development requires strategic efforts involving the entire organization. Maintenance efforts also play an important role. Company management needs to understand and develop an appropriate strategy to achieve sustainable development by applying maintenance performance measurements. The aim of this paper is to present possible ways of analyzing and ranking the impact of certain criteria with respect to achieving sustainability. The paper uses the method of Structural Equation Modeling—SEM in order to determine the most influential variable on the sustainability of maintenance of technical systems. Based on the set theoretical system model, for all its variables in the model, statements were made that describe them, on which 136 respondents gave their views (from 1 to 5, Likert scale) in the territory of the Republic of Serbia. An intuitive F-DEMATEL method was also used to prioritize variables. A team of 10 experts in the field of maintenance of technical systems was compared the criteria A—Application of technical diagnostics, B—Management of maintenance resources, C—Maintenance process planning, and the dependent variable D—Sustainability of maintenance of technical systems. According to experts, the importance of the criteria coincides with the results obtained by a survey with 136 respondents.

In this work, we show the experience of the driver rehabilitation process in the Republic of Serbia, with the analysis of the rehabilitation process and the changing of the drivers’ attitudes. Before performing the analysis, we define the basic hypothesis (implementation of the sustainability model of E learning for the driver rehabilitation process will impact the driver attitude and minimize the mistakes). In the analysis, we take the most recognized mistakes in the process of education. Implementation of this model is exclusive for a group of candidates who finished the driver rehabilitation process. Using the database of drivers who made mistakes, we made the electronic database (video material with comments) and subsequently implemented the driver education model to assess who made mistakes in the rehabilitation process. In the test process, we used the two groups of candidates (age 18–30 and 31–40), before and after the implementation of this e-learning model. In order to process the data, taking into account the small number of observations further in the research, correlation analysis was used, which determines whether there is a correlation between the amount of alcohol in blood (expressed in g/kg) and the reaction time (in seconds) of a driver. In the second part of the work, correlation analysis was performed with the aim to examine which type of relation exists between the candidates who participated in the driver rehabilitation program and candidates that participated in the program of the proposed e-learning model. The results of this study offer a direction in which learning takes place by itself and affects the effective environment of e-learning, multimedia teaching, virtual practical content, and learning under the instruction of the lecturer in this target group of participants in the rehabilitation program. How many times have we believed that we have an ability and overestimated our-selves in that? You can be convinced in traffic only if you have to prove that ability. The results of the application of this model make it possible to prove that ability and change the opinion about the stated mistakes. The key goal of this work is to make the sustainability e-learning model based on mistakes, which will eliminate the same mistakes made by drivers. The result would be top training and high-quality performance in traffic.