CDNA - A Context-Aware Notification System for Driver Interruption

Inopportune driver notifications are a real problem that may cause distractions and interruptions in traffic, and hence accidents. Notifications on mobile devices are one of the ways by which drivers are interrupted, reaching extremely high amounts in a normal day. Despite that, notifications are valued by users and they are part of the common use of smartphones. Therefore, how to lessen the interruptive potential of notifications without eliminating them completely? To mitigate this problem, the present work proposes a context-aware notification system with the identification of opportune and inopportune moments for drivers notification. The proposed system uses smartphone sensors (gyroscope and GPS) to infer if the driver may be interrupted in a specific moment to receive a notification. Preliminary experiments were performed with people in real driving situations to verify if the system could identify opportune and inopportune moments, and found results indicate that is possible to identify these moments with a general accuracy of 88%.