Automated passengers trip phase recognition and public transit accessibility level analysis via machine learning models using GPS data

Determining true value of passengers waiting time, walking time and distance to and from public transit stops from passengers and public transit side is a key index to assessing potential demand, quality, and effectiveness of public transit services. Current studies heavily rely on household surveys, direct observation techniques, GIS, and telephone-based interviews. However, some investigations use GPS trajectories as a primary data source to detect trip phases, and the main drawback is saving a few trips. This study as a first investigation in this field tries to detect trip phases to and from public transit stops for both passengers and infrastructure side with an automated trip phase recognition algorithm. Passengers waiting time as a critical feature for public transit planning also infer from raw GPS data to analyze the performance of bus stations. Moreover, for bus stations in public transit-based trips the distribution of waiting time, access and egress time, and distance where there is GPS data have been computed. A random forest model, extract transit modes for each segment of a trip, and the results will be used to detect trip phases. There is a lack of labeled standing data in Geolife dataset to detect waiting time. Another novelty is to combine Sussex and Geolife as two large datasets to increase number of estimated modes, especially standing. Our results underscore the effectiveness of our automated approach in predicting different phases.