Model Driven Software Development for Human-Machine Interaction Systems

In a typical Human-Machine Interaction (HMI) system, a task is performed by cooperation of the human and the automation component. The system adopts a cognitive architecture to model human psychology and makes optimum decisions on dynamic task allocation between human and the machine counterpart depending on the context. However, such architectures do not define how those systems are implemented in software. Various models involved in Model Driven Software Development (MDSD) approach in developing HMI systems is presented. This paper proposes a metamodel for modeling Non-Functional Properties (NFP) in HMI systems and provides a case study on assistive lane keeping in automobiles to demonstrate the approach.