Opportunistic IoT Service to Support Safety Driving from Heterogeneous Data Sources

The Internet of Things (IoT) represents an ecosystem where heterogeneous components seamlessly interoperate aiming to provide opportunistic (highly contextualized, dynamic, transient, and co-located) cyberphysical services in every application scenario, including smart automotive. Just in the context of advanced driving assistance systems, this paper proposes a modeling approach supporting the interactions among multiple Smart Objects (SOs, like Smartphone, Smart Bracelet, Smart Cushion, etc.) within the vehicle, in order to retrieve information regarding driver psycho-physical status and to alert if risky conditions (i.e., distraction, drowsiness, high stress level, or aggressive behaviors) are detected. The outlined “Driving Assistance Service” is expected to collect contextualized data from heterogeneous SOs (not purposely designed for implementing such service nor for interoperating), to perform data fusion and analysis, and finally to provide multimodal alerts on the driver’s smartphone. The goal of this work, indeed, is to show that the proposed metamodel-based approach facilitates the implementation of such an integrated IoT service, also improving the embedded and closed ADASs currently available at the state-of-the-art.