On quantitative assessment of reliability and energy consumption indicators in railway systems

Stochastic model-based approaches are widely used for obtaining quantitative non-functional indicators of the analysed systems, as for example reliability, performance and energy consumption. However, a critical issue with models is their validation, in order to justifiably put reliance on the analysis results they provide. In this paper, we address cross-validation on a case study from the railway domain, by modelling and evaluating it with different formalisms and tools. Stochastic Activity Networks models and Stochastic Hybrid Automata models of rail road switch heaters, developed for the purpose of evaluating energy consumption and reliability indicators, will be evaluated with Mobius and Uppaal SMC. We will compare the obtained results, to improve their trustworthiness and to provide insights on the design and analysis of energy-saving cyber-physical systems.