Skip to Main Content
This paper addresses variability in the domain of software-based control systems. When designing product lines of such systems, varying sensors and actuators have to be used and parameterized, which in turn requires adaptations in the behavior of the microcontroller. For efficient engineering these adaptations should be performed in an systematic and straightforward manner. We tackle these challenges by using a Rapid Control Prototyping (RCP) system in combination with model-based development techniques. In particular, we modularize the parametrization of components into a separate configuration, which is isolated from the model that defines the controller behavior. Hence, during adaptations the model can often remain unchanged, which significantly reduces the turnaround time during design iterations. The approach is illustrated and evaluated with a parking assistant application, which is tested on our experimental vehicle, where it performs automatic parking maneuvers.