360

Analysing Google Street View images allows us to account for static road factors influencing safety - infrastructure - and average measures of dynamic factors (pedestrian and car density). Although, it is also highly relevant the detection in real-time of vehicles and people because in most accidents they are involved.

By incorporating, in scooters and bicycles, cameras able to capture the surroundings 360º and in real-time, our assisted-driving model is able to detect with higher than 95% accuracy vehicles most often involved in crashes. This will allow us to perform actions (varying vehicles’ speed and provide luminous/sound feedback to the rider) intended to prevent collisions.