Crowd-Based Road Surface Assessment Using Smartphones on Bicycles

Abstract

The surface quality assessment procedures of roads used by motor vehicles are widely addressed in research, recently with a focus on continuous assessment using crowd data. Although, the bicycle infrastructure is becoming increasingly important for urban mobility and the shift towards sustainable transportation, its monitoring and maintenance is currently underrepresented in research and in the awareness of road authorities. In this paper, we evaluate different approaches to assess the surface quality of bicycle lanes and paths using crowd data from smartphones carried on bicycles. The applicability of data from smartphone sensors acquired on moving bicycles is quite limited in its usage towards the calculation of established surface assessment metrics. Thus, we consider various metrics, with a focus on robustness regarding the named limitations. The evaluation is done in a first step in controlled test rides on known surface types. In the second step, the most promising metric is evaluated against its reproducibility with different bikes and riders on multiple test rides on a random track.