Multi-Sensor Architecture in a Prototype Vehicle Within a Smart Road Scenario for Driver, Vehicle, and Surrounding Environment Monitoring Based on Wired/Wireless Solutions

Urban traffic is increasingly complex, and even minor errors can lead to tragedies. Every day, thousands of vehicles, pedestrians, and cyclists share a dynamic urban space, posing significant risks. Accidents are often caused by reckless driving, but also by structural limitations such as unpredictable road conditions and incomplete perception of vehicle surroundings. In this context, an innovative solution has been designed to improve driver safety through a multisensory approach that monitors the driver, vehicle, and external environment. The on-board system relies on perception sensors to detect asphalt conditions, obstacles, and tire wear, along with a Driver Monitoring System (DMS) that acquires inertial, postural, and biophysical data (ECG, EDA) through wireless modules. This article presents the design and testing of steering wheel and seat belt modules, including block diagrams, custom electronic sections for driver parameter acquisition, 3D-printed cases, and seat cover design. It also describes the wired and wireless architectures that enable the NVIDIA Jetson board to acquire driver, vehicle, and surrounding-environment information and run ML/DL algorithms to detect critical driver (drowsiness, illness, drunkenness) and environmental (obstacle, black ice, and pothole) conditions. Steering wheel cover characterization against reference instruments yielded a HR MAE of 2.1 bpm (r = 0.96) and skin conductance MAE of 0.9 µS (r = 0.89), while the photovoltaic section harvested 360, 180, and 60 mW in direct sun, penumbra, and indoor lighting conditions.