Research Summary: SafeStudent Driving: A Multimodal Driver-Safety System to Support Teen Drivers Using Computer Vision and Mobile Sensing

A multimodal driver-safety system, SafeStudent Driving, has been developed to address high crash rates among teen drivers. This system integrates computer vision for detecting traffic lights, road signs, and speed limits, alongside mobile sensing for inferring turn signal usage. The platform utilizes both a Raspberry Pi and a Flutter-based mobile application to deliver prioritized voice prompts, aiming to improve driving behavior through real-time feedback. Key challenges in development included maintaining model accuracy across diverse lighting conditions and optimizing inference for efficiency.

Why it matters

This development introduces a novel application of AI and mobile technology to enhance road safety, specifically targeting a demographic with elevated risk. Successful implementation and broader adoption could lead to a reduction in accident rates, lower insurance costs, and improved public safety outcomes, thereby influencing policy and technology integration in transportation sectors.

Key insights

• Teen drivers exhibit a disproportionately high rate of traffic accidents, primarily attributed to inexperience and inconsistent attention.
• The SafeStudent Driving system employs a multimodal approach, combining computer vision and mobile sensing technologies.
• Computer vision components include three YOLO-based models for traffic lights, light-bulb colors, and road signs, and an OCR module for speed limit detection.
• Mobile sensing integrates an audio model and IMU data to infer turn signal usage.
• The system operates on a Raspberry Pi and a Flutter-based mobile application.
• Real-time feedback is provided through prioritized voice prompts, generated via text-to-speech or pre-recorded audio.
• Significant development hurdles involved achieving consistent model accuracy in varied environmental lighting and optimizing inference processes.

Source

arXiv — Computers and Society — https://arxiv.org/abs/2608.07487