Stochastic index definition and estimation for reliability and quality assessment of transportation systems

Abstract

In the paper some considerations are performed with respect to reliability and quality methodologies applied to transportation systems. These systems have to be considered as large complex systems in which the two aspects are closely related, and which experience fast variations in time, so that a satisfactory description must not disregard dynamic aspects. According to this philosophy, the authors propose an index, also directly oriented to customer satisfaction, which in a global way measures the quality of the system (here meant as “service” quality). This index is defined in a probabilistic way in terms of delay with respect to the scheduled timetable, by properly linking this aspect to the faults occurring in the system under investigation. Such approach allows to connect service quality with system reliability. The advantages of the proposed index are discussed and a numerical application is reported in order to demonstrate the main features of the proposed index, which can be expressed analytically under reasonable hypotheses, and also coincides in one case with an Extreme Value distribution function. Finally, hints at possible adoption of a Bayesian method for its estimation are provided. In the course of the paper, some misconception are highlighted which exist in literature about the interpretation and evaluation of common performance indexes such as failure (or hazard) rate and failure intensity.