Design environments for complex systems

The paper describes an approach for modeling complex systems by hiding as much formal details as possible from the user, still allowing verification and simulation of the model. The interface is based on UML to make the environment available to the largest audience. To carry out analysis, verification and simulation we automatically extract process algebras specifications from UML models. The results of the analysis is then reflected back in the UML model by annotating diagrams. The formal model includes stochastic information to handle quantitative parameters. We present here the stochastic π-calculus and we discuss the implementation of its probabilistic support that allows simulation of processes. We exploit the benefits of our approach in two applicative domains: global computing and systems biology.