This article presents a study on the benefits offered by Coloured Petri Nets in capturing and separating permanent and temporary behavioural information and on the systematic support they hereby provide to model-based design and testing of cyber-physical systems. In particular, it illustrates the application of CPN modelling to capture the behaviour of cooperative mobile robots and highlights their benefits in terms of compactness and scalability. Finally, the article reports on the applicability of test case generation algorithms supporting the coverage of the underlying CPN models with respect to different testing criteria.