Abstract
In the last few years, with the exponential diffusion of smartphones, services for turn-by-turn navigation have seen a surge in popularity. Current solutions available in the market allow the user to select via an interface the desired transportation mode, for which an optimal route is then computed. Automatically recognizing the transportation system that the user is travelling by allows to dynamically control, and consequently update, the route proposed to the user. Such a dynamic approach is an enabling technology for multi-modal transportation planners, in which the optimal path and its associated transportation solutions are updated in real-time based on data coming from (i) distributed sensors (e.g., smart traffic lights, road congestion sensors, etc.); (ii) service providers (e.g., car-sharing availability, bus waiting time, etc.); and (iii) the user’s own device, in compliance with the development of smart cities envisaged by the 5G architecture. In this paper, we present a series of Machine Learning approaches for real-time Transportation Mode Recognition and we report their performance difference in our field tests. Several Machine Learning-based classifiers, including Deep Neural Networks, built on both statistical feature extraction and raw data analysis are presented and compared in this paper; the result analysis also highlights which features are proven to be the most informative ones for the classification.
Highlights
This paper presents a series of Machine Learning-based classifiers for real-time solution of the Transportation Mode Recognition (TMR) problem, i.e., the correct classification of which transportation system the user is travelling with
This section describes three TMR solutions based on statistical feature extraction, i.e., Random Forests, Feed-Forward Neural Networks and Recurrent Neural Networks, and two solutions based on raw data, i.e., a deep implementation of Feed-Forward Neural Networks and a Deep
The results of the field tests the section, are in line with the results of the tests performed on the validation sets; validation, reported are in line with the results of the tests performed on the as an example of confusion matrices, we report at the end of the section only the ones relative to the validation sets; as an example of confusion matrices, we report at the end of the section
Summary
This paper presents a series of Machine Learning-based classifiers for real-time solution of the Transportation Mode Recognition (TMR) problem, i.e., the correct classification of which transportation system the user is travelling with. The statistical feature extraction-based approaches will be trained and tested on data gathered during a time window of 2 s, while the raw data analysis solutions require approximately one second of samples, without any significant computational overhead As it will be discussed, the main characteristics of this paper with respect to the state of the art are the following:. This analysis is often conducted on statistical data obtained from socio-economic factors [4], but the availability of a real-time information flows enables the optimization of the aggregated transportation flows, by means of a dynamic response to specific demands (e.g., increasing the number of busses during certain periods of time over the day/year, or re-planning their routes) This approach would lead to a tangible increase in public transport users’ experience and, as a by-product, to a significant increase in the cities’ welfare and to the attraction of more investments.
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
