The integration between Artificial Intelligence and smartphone sensors is defining the future of Smart Cities. A series of studies led by researcher Carlos Quintella, creator of the CityTracks app, demonstrated that it is possible not only to identify how citizens travel (bus, car, on foot), but also the purpose of the trip with high precision.
The highlight of the research is the combined solution published in IEEE Transactions on Intelligent Transportation Systems, which uses a unique preprocessing algorithm to extract mobility patterns. By applying Automated Machine Learning, the system achieved 95% accuracy in detecting the means of transport and according to the researcher, it will soon be able to identify the reason for the trip (e.g. work, leisure). <a href="https://ieeexplore.ieee.org/document/7795921">Detecting the transportation mode for context-aware systems using smartphones</a>
This paper presents a classification method for smartphone users mobility data in urban environments according to the used transportation mode. This classification is possible among several different transportation modes and using only the location data from user's mobility. Among the methods applied, includes data mining with machine learning techniques for the inference. This paper also presents: the performance analysis for several machine-learning algorithms for the proposed task; the process used to collect mobility data for nine users along six months; the process used for data pre-processing and the computational architecture used to collect participatory sensing data.