Real-Time Driver Identification System in the Internet of Vehicles: Seamless Integration of Deep Learning and Cloud Computing

The Internet of Vehicles (IoV) represents a transformative extension of the Internet of Things (IoT), with a specific focus on revolutionizing the automotive industry. In this dynamic landscape, the importance of developing a seamless and integrated system cannot be overstated, as it plays a crucial role in enhancing operational efficiency and ensuring safety. This research introduces an innovative approach for real-time driver identification detection by harnessing the combined power of deep learning's robust classification capabilities and the virtually limitless resources offered by cloud computing. By leveraging the capabilities of Google Cloud, Thingsboard, and the real-time communication technique of pub/sub, the developed solution is tailored specifically for IoV technology, effectively managing real-time data collection, processing, prediction, and visualization while maintaining resilience against anomalies in sensor data. Additionally, this study proposes a hybrid deep learning approach that integrates LSTM and multi-head self-attention mechanisms. The proposed model is rigorously validated using two datasets, including Security and collected datasets, demonstrating its superiority over previous methods with remarkable accuracy and F1 scores of 99.7 and 99.47, respectively. By achieving precise driver identification outcomes, the proposed end-to-end IoV system holds significant potential in optimizing fleet management, enhancing vehicle security, personalizing driving experiences, enabling effective insurance and risk assessment, and ultimately contributing to road safety and efficient transportation management.