A monitoring framework for urban road ride quality using smartphone sensing technology

•The vibration measurement capability of the smartphone was investigated through shaking table tests.•A scalable deep learning was utilized to quantify the driving comfort intensity from whole-body vibrations.•A multi-source driving comfort data fusion strategy was proposed to monitor the road sections with poor ride quality.•A systematic deployment proposal for monitoring the ride quality of urban roads was introduced using the crowdsourced spatiotemporal data. Ride quality monitoring is crucial for the sustainable development of modern urban transportation infrastructure. This study innovatively designed a two-phase synergetic framework for monitoring the ride quality of urban roads using smartphone sensing technology. Firstly, a convolutional neural network integrated multi-head attention mechanism was developed to quantify the driving comfort from whole-body vibrations. Then, a multi-source driving comfort data fusion strategy was proposed to monitor the road sections with poor ride quality. The experimental results show that: (1) the consistency between the developed model outputs and the results quantified by ISO 2631–1 exceeds 90%; (2) visual inspection results of ride quality monitoring align with actual pavement conditions. Compared to existing studies, the designed framework avoids over-reliance on prior knowledge and improves objectivity and reliability of ride quality monitoring. This study provides a novel insight into urban road monitoring and thus contributes to the refined management of urban road assets.