Modeling the MVM-Adapt System by Compositional I/O Abstract State Machines

With the increasing complexity and scale of software-intensive systems, model-based system development requires composable system models and composition operators. In line with such a vision, this paper describes our experience in modeling the behavior of the MVM-Adapt, an adaptive version of the Mechanical Ventilator Milano that has been designed, certified, and deployed during the COVID-19 pandemic for treating pneumonia. To keep the complexity of the requirements and models under control, we exploited a compositional modeling technique for discrete-event systems based on Abstract State Machines (ASMs). Essentially, separate ASMs represent the behavior of interacting subsystems of the MVM with their new adaptive functionalities; they can communicate with each other through I/O events, and co-operate by a precise orchestration schema.