Edge system predicts cyclist safety scores from handlebar cameras
Edge-Assisted Crowdsensing for Explainable Cyclist Safety Perception
Computer Vision and Pattern Recognition
Summary
Cycling can be dangerous, and it’s hard to continuously monitor how safe cyclists feel on their routes. The authors made a system that uses cameras mounted on bike handlebars to capture street views, and a special AI model to understand what’s in each picture. By combining this information with map data and bike movement, their system estimates how safe cyclists feel along the way. This helps city planners see which road features make cycling safer or more risky, supporting better decisions for cycling infrastructure.
What this means in practice
- •For urban planning teams: Use continuous cyclist safety scores from real trips to prioritize road infrastructure investments that improve safety based on explainable model insights.
- •For transport network operators: Monitor real-time cyclist safety levels over city routes at scale to optimize traffic management and reduce accident risks involving cyclists.
Authors
Hongsen Cao, Chia-Yen Chiang, Mona Jaber, Meihui Wang, Yasmin Fathy
Abstract
Cyclist fatality rates remain high despite policy support for active travel, and autonomous vehicles add complexity to mixed traffic. Neither crash statistics nor multi-model detection pipelines scale to continuous monitoring of perceived safety along a route. We present an edge-assisted crowdsensing system: handlebar cameras supply street-level frames, a vision-language model (VLM) turns each frame into structured traffic attributes through a single prompt, and an explainable boosting machine (EBM) combines them with OpenStreetMap road infrastructure and on-bike kinematics to predict a crowdsourced 1-9 safety score. Extraction belongs at the network edge: the fastest self-hosted model needs 2.49 s per frame and scoring takes 0.4 ms on a CPU. On 1,362 annotated frames from two London routes, the EBM regressor reaches 72.2 percent binary accuracy and a mean absolute error of 1.011 under five-fold cross-validation. It is within one accuracy point of black-box ensembles. Its per-feature shape functions give exact attribution: physical lane separation is worth +0.97 rating points, a painted lane +0.70 and a construction site -0.65. Planners can therefore price an infrastructure change in rating points, and the results enter cost-benefit analysis directly. We also report inter-annotator agreement (alpha = 0.4695), an attribute audit, and a sensitivity analysis linking attribute error to accuracy. With re-encoding, a 20 s sampling interval serves 1,000 riders at 54 Mbit/s at a cost of 1.4 accuracy points, making continuous monitoring practical.