We provide a brief comparison of the modelling and analysis capabilities of two different formalisms and their associated simulation-based tools, acquired from experimenting with these methods and tools on one specific case study. The case study is a cyber-physical system from an industrial railway project, namely a railroad switch heater, and the quantitative properties concern energy consumption and reliability. We modelled and analysed the case study with stochastic activity networks and Möbius on the one hand and with stochastic hybrid automata and Uppaal SMC on the other hand. We give an overview of the performed experiments and highlight specific features of the two methodologies. This yields some pointers for future research and improvements.