Work

Smartphone based applications for Road Traffic Telematics

Year: 2019
Type: dissertation
Abstract: This research is concerned with Intelligent Transportation System (ITS), with two major points of focus.The first point is the collection and processing of data from road traffic using smartphones and other devices (such as On-Board Units (OBUs)).It aims at using smartphones that are ubiquitous and host various sensors whilst providing several communication interfaces, to collect data and to investigate the possible beneficial applications of such data for traffic management, awareness and safety.The collected data is of high value for relevant authorities such as city management, public transportation system, traffic police, vehicle insurance companies, etc.The second point is the use of machine learning techniques to predict the intensity of traffic using crowd-sourced data from a small segment of the traffic.The processed information about traffic intensity can improve the accuracy of Advanced Traffic Management System (ATMS) and reduce the costs incurred due to the use of dedicated traffic sensing hardware.A number of facts motivated this work, some of which are as follows.First, the high demand for mobility: whereby more than 70% of all journeys are made by a car in the European Union (EU).Specifically in Italy, where the number of vehicles per 1000 inhabitants is 625, which places it the 2 nd highest across the EU.Second, the cost of traffic congestion: one consequence of high motorisation rates is traffic congestion, which costs about € 100 billion in the EU every year.Third, the growing adoption rates of smartphones: smartphones are omnipresent and well-connected; almost 80% of Internet users in the EU surfed via a mobile device in 2016; the average Penetration Rate (PR) of smartphones in the EU is quite high at about 67.3% of its population.Fourth, the versatility of high-resolution and high-quality mobile sensor data: smartphones have high computational power, high capacity connectivity, and low-power Inertial Measurement Unit (IMU) which enable them to be orientation-aware with minimal power consumption.Nearly every single smartphone produced in the last decade has a 6 or 9-axis IMU built-in.And finally, cloud-based data crowdsourcing trend: thanks to affordable Internet connectivity, cloud-based crowdsourcing for data has opened new doors for providing rich services to users without any dedicated hardware for data collection.To further elaborate on the first point, ITS applications that require data related to vehicle dynamics (e.g.acceleration, yaw angle, etc.) usually have low PR iii 7.24 Training dataset size comparison shown by box plot grouped by training PR. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .A.1 AS dataset CRS test results. . . . . . . . . . . . . . . . . . . . . . .A.2 DT dataset CRS test results. . . . . . . . . . . . . . . . . . . . . . .A.3 AS&DT dataset CRS test results. . . . . . . . . . . . . . . . . . . .A.4 AS&DT&H dataset CRS test results. . . . . . . . . . . . . . . . . .A.5 IA&H dataset CRS test results. . . . . . . . . . . . . . . . . . . . .A.6 IA_2CH dataset CRS test results box plot grouped by binning window size. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .A.7 M_AS&DT dataset CRS test results. . . . . . . . . . . . . . . . . .A.8 IA 5-categories dataset CRS test results box plot grouped by training PR. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .A.9 IA 5-categories dataset CRS test results box plot grouped by testing PR. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .A.10 IA 8-categories dataset CPR test results box plot grouped by training PR. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .A.11 IA 8-categories dataset CPR test results box plot grouped by testing PR. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .A.12 IA 8-categories DPR dataset CPR test results box plot testing only with 1% PR. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .A.13 IA 8-categories DPR dataset CPR test results box plot testing only with 5% PR. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .A.14 IA 8-categories derived and variable PR dataset CPR test results box plot testing only with 1-2.5-5-7.5-10VPR. . . . . . . . . . . . .xv Mobility demand and traffic congestionPeople are on the move more than ever in the world, but particularly in the European Union (EU).In the EU, more than 70% of all journeys are made by a car (be it a private car, a taxi or a car-sharing service).It is evident that mobility is becoming more and more essential and vital due to ever-increasing distances.Due to the growing urban population and consequently growing urban cities, distances between home, work, schools/colleges/universities, and other facilities are Data crowdsourcingCrowdsourcing is the practice of obtaining information or input into a task or project by enlisting the services of a large number of people, either paid or unpaid, typically via the Internet according to the Oxford dictionary.Cloud-based crowdsourcing for data has opened new doors for providing rich services to users without any dedicated hardware for data collection.This approach consists of building large data sets (real-time or not) with the help of a large group of people using a medium such as a smartphone application.This phenomenon is of paramount importance in
Source: PORTO@Iris
Language: English
Subfield: Transportation
Open Access status: green
OpenAlex ID: W2989724167
Cites: 44
Cited by:
Related to: 10