Description

As cycling population increases, exposure to bicycle crashes also increases, which opens the question 'Why do we crash on the bicycle?'. Currently we don't have a conclusive answer. Common factors involved in crashes are known, such as slippery conditions, external perturbations and avoidance manoeuvres among others. However, the aim now is to delve into human biomechanics to understand why the rider is unable to resolve the critical situation and ends in a fall.

This project consists in obtaining real-world bicycle crash data, create simulations that reproduce crash scenarios and include relevant human biomechanics for the outcome. With this, we aim to contribute to the understanding of human behaviour in critical situations, which will improve road safety.

Ongoing Work

• Single-cyclist crashes classification: We know the hypothesised mechanisms that are present in bicycle crashes; however, they have not been proven or discussed based on bicycle dynamics theory. The goal is to discuss these mechanisms with the theory and provide a concise background for classifying bicycle crashes. Therefore, we provide a classification that takes into account different layers that contribute to the outcome of the critical scenario. In addition, we consider how the biomechanics and control of the rider play a role in the mechanisms to highlight the knowledge gap in this area.
• Kinematic data from monocular video: Real-world data of single bicycle crashes is scarcely available; therefore, we aim to extract the kinematics of the system from monocular videos. To this end, we are using computer vision methods such as instance segmentation to track the motion of the bicycle in the scene.

Contact

Do you want to learn more about this project, collaboration options or themes for MSc and BSc Thesis? Reach me out!

References