A queue analytics system for taxi service using mobile crowd sensing
Journal Title:
Adjunct Proceedings of the 2015 ACM International Joint Conference on Pervasive and Ubiquitous Computing and Proceedings of the 2015 ACM International Symposium on Wearable Computers
Passengers waiting queues and taxis waiting queues are commonly seen in many urban cities. Our poster presents a queue analytics system, which collaboratively uses the mobile data from taxis and smartphones, to detect both passenger queues and taxi queues. In particular, the system firstly determines the existence of taxi queues by analyzing the taxi data, and then make a soft inference on passenger queues. Meanwhile, the passenger side adopts the smartphone-based crowd sensing strategy to detect the personal-scale queuing activities. Lastly, the system aggregates the detection results and validates passenger queues. The extensive empirical experiments demonstrate our system can accurately and effectively achieve the design objectives. Moreover, the system envisions a novel crowd sensing way to perform online analysis using data from heterogeneous sources.