We presented the work conducted in the FSF project to han-dle mixed criticality. We used a synchronous design frameworkto implement a simplified signaling application and to deployit on a partitioned OS.We are continuously working towards a better integrationof the tools composing the framework.In the passenger exchange use case, mixed criticality residesat the application level, or even at function level, rather thanthe system level. On the other hand, the approach proposedin IMA and ARINC meets the needs of a system integrator.The main constraint highlighted by this case study is that theremay be, even within a single system function, many communi-cations between its vital and non-vital subcomponents. Whengeneralized to the whole set of system functions, this patterninduces a large number of communications between the vitaland non-vital parts. Furthermore, if we want to preserve thesynchronous semantics (e.g. no additional delay) the numberof windows may explode. The overall cost of communicationsand context-switch become prohibitive for systems global per-formance. Executing mixed-critical signaling applications onthe same platform remains a challenging problem consideringthe state of the art in real-time operating systems.Finally, the vital/non-vital dichotomy traditionally used insignaling application proved to be insufficient with respectto the operational availability of the system. It would bemore appropriate to consider at least three levels, safety-critical, mission-critical, and non-critical, and to exploit thisinformation in the partitioning and scheduling.