CLACSOON: Carpooling in Urban Areas

The research presented in this thesis focuses on ICT technologies for urban mobility, with particular attention to the study, design and the development of applications for dynamic carpooling services in smart cities. The work has been done within the PhD programme in apprenticeship, during the development of the CLACSOON project while working for the startup company GreenShare SRL. Nowadays, the development of a Smart City represents an approved way to improve the quality of life in the urban context. The sustainable transport is a key service for a smart city, and carpooling solutions have gained more and more popularity in the last years. This thesis mainly focuses on the challenges of a real-time carpooling service. A lot of effort has been made on the design and the development of a Cloud-Mobile Dynamic Ridesharing platform named “CLACSOON”. The proposed solution is an application that automatizes the arrangement of the shared ride, automatically notifying the presence of suitable travel companions and suggesting the pick-up points and the meeting times. The pick-up and drop-off point are computed by a route matching algorithm that aim to introduce two novel features for a ridesharing application. The former is the partial ridesharing, according to which the riders can walk to reach the driver along his/her route when driving to the destination. The latter consists in the possibility to share the ride when the driver has already started the ride by modeling the mobility to reach the driver destination. To evaluate performances of the service and the Quality of Experience (QoE) provided to the users, an emulation system has been implemented to evaluate the performance of the CLACSOON platform. The performances, expressed in terms of Key Performance Indicators (KPI), show that introducing these features in a route matching algorithm leads to a substantial performances improvement; moreover, the results can be considered to evaluate the requirements to build a successful urban carpooling service. Additionally, the carpoolers’ cars, with the coming of the IoT, could be seen as a formidable sensor platform. Since one of the aspects of smart cities is the optimal use of the resources, the information coming from these sensors can be processed and analyzed to improve the efficiency and quality of the urban services and to support smart transportation and sustainable mobility. In this context, Machine to Machine (M2M) and Device to Device (D2D) communications play an important role. Starting from this vision, a distributed algorithm for the task allocation and assignment, which can be executed by a group of real IoT devices, has been developed with the aim of maximizing the lifetime of groups of nodes involved while ensuring the fulfillment of the requested Quality of Information (QoI) requirement. In the field of infomobility and support of the urban mobility, it could be run in real devices on the vehicles.