Data crowdsourcing and traffic sensitive routing for a mixed mode public transit system

In this paper, we propose transit journaling, a crowdsourcing solution for public transit data collection, and we describe CommYouTer, an Android app for this purpose. CommYouTer enables the user to (1) document his public transit trips with automated transfer detection, (2) participate in crowdsourcing real-time traffic conditions and other relevant data, and (3) query for efficient commuting directions via a traffic-sensitive routing algorithm. For transit journaling, the app offers a recording feature that uses the smartphone´s GPS antenna and accelerometer to track the user´s location and activity. Activity detection is applied to the mobile phone´s accelerometer data to differentiate between two user states (walking vs. non-walking), which are then used to determine vehicle transfers along the journey. We also implement our modification of RAPTOR, an existing round-based public transit routing algorithm. Our modification allows the system to account for real-time crowdsourced traffic conditions. We test our system in Metro Manila, Philippines, where public transit is primarily headway-based (non-scheduled).