Regresar

A methodology for train trip identification in mobility campaigns based on smart-phones

Abstract:

Nowadays, mobility campaigns use mobile phones as sensors for travel surveys aimed at gathering chronological information, patterns and modes used by citizens. Train trip travel identification is one of the issues present in this new schema. Differentiating train and car trips is challenging because in many cases railways and roads are side by side and their individual travels have similar speed. In this paper, we describe a methodology based on a speed-based filter and geospatial operation using the OSM network to determine possible train trip segments in data gathered in a mobility campaign. We evaluated our method using over 9,683 segments, which have been gathered by 239 devices. The results show that the proposed approach successfully detects 76.14 of the train trip segments labeled by users. This methodology can be used as a post-processing step to classify train segments in big data of smart cities.