The fleet and breakdown data that is provided by the user is quickly cleaned up in several steps and prepared in such a way that it can be safely processed further. Not only invalid and duplicate values are removed, but also non-monotonic entries. The process has been made more resilient and automated with many datasets of operational fleets.

An optimal model structure is automatically sought for the available failure data in order to determine the actual number of failure components. Various model candidates are trained with the Expectation Maximisation Algorithm and compared with each other. The optimum structure is determined on the basis of statistical criteria so that no manual intervention is necessary. This makes the application possible even for non-experts.

The lifetime forecast models the milage of each vehicle individually. The resulting probabilistic ageing models depict future ageing as a probability distribution. This makes it possible to estimate the mileage distribution of a fleet much more accurately than with the widely used three-point estimate.

The method is based on the consistent use of probabilistic models, both for mileage estimation and for the probability of failure. The underlying uncertainties in the data are taken into account within both models and offset against each other. As a result, the method provides not just a single point estimate, but the entire probability of occurrence.

When looking at the service life curve alone, it is only possible to make a generalised statement about the number of failures at a certain mileage. In reality, however, vehicles drive differently and there is no fleet in which all vehicles have the same mileage. To predict the number of failures, the probabilistic service life forecast therefore considers the vehicle-specific mileage and forecasts this into the future in order to draw conclusions about future failures from vehicle-specific failure risks.