Przemysław Wrona
supervisor: Maciej Grzenda
The proliferation of various sensor technologies in smart cities, the prevalence of mobile phones, the internet-of-things technology, and networks of sensors has led to an enormous and ever-increasing amount of data that are now more commonly available in a streaming fashion. One type of never-ending stream is sent current localization of public transport such as busses, trams or metro lines.
We present the system for detecting delays in public transport using concept drift detection methods. System integrating raw data from Warsaw Public Transport coming from the data stream with timetables. Based on vehicle arrivals from streams and schedules, there will be calculated delay, which in the future will be used to build a model for predicting delay.
The approach used in the work is the adaptation of traditional machine learning algorithms and artificial neural networks for a non-stationary environment, in which the data for training the model flow in an endless stream.
This work investigates day-to-day variability in public transport travel time using a GPS data set for public transport routes. It explores the nature and shape of travel time distributions for different departure time windows at different times of the day and factors causing travel time variabilities of public transport, such as distance between stops and destination, quality of vehicle number of seats or delay at the previous stop. We demonstrate the system with a real-world use-case at Warsaw city, Polan