The Rider Awareness System, (Project RAS) aims to reduce cyclist and motorcyclist fatalities by providing the rider with high situational awareness. Using a YOLO-powered detection system, it warns the rider of nearby vehicles, provides real time information to riders through their phones to help avoid collisions, and suggests evasive maneuvers.

Project RAS integrates computer vision, using the YOLOv4.lite model for real-time object detection to identify approaching vehicles. The hardware includes a rugged, customized Makita case attached to the rear bike rack, containing a laptop, PS3 Camera, USB hub, and power bank, and speaker at front, which are adaptable for rugged bike use. The system’s software is developed off python for object detection, while OpenCV and YOLO provide real-time image processing. The powerbank powers an attached light strip, improving visibility to other vehicles. The speaker and phone make it possible for the rider to receive visual and auditory warnings.

Project RAS demonstrated reliable detection of approaching vehicles in tests. However, limitations included occasional overlaps in readings and sensitivity to low-light conditions. The project successfully met its objectives, providing timely and accurate alerts for rider awareness, though further testing and iterations are required to improve the robustness of project RAS.

Project RAS demonstrates the potential of ML models (YOLO) for rider safety and suggests these systems should be standardized on NSW roads. It highlights the value of rider situational awareness, with room for improvement in low-light conditions to ensure effectiveness across all environments.