Authors:
Milton Boos Junior
1
;
Lucas Sakurada
1
;
Paulo Leitão
1
;
Paulo Alves
1
;
Gleifer Vaz Alves
2
;
André Pinz Borges
2
and
Diego Roberto Antunes
2
Affiliations:
1
Research Centre in Digitalization and Intelligent Robotics (CeDRI), Instituto Politécnico de Bragança, Bragança, Portugal
;
2
Federal University of Technology - Paraná (UTFPR), Campus Ponta Grossa, Paraná, Brazil
Keyword(s):
Multi-Agent System, Cloud Architecture, Smart Parking.
Abstract:
Smart parking (SP) systems are becoming a solution to address the increasing traffic in major cities, which are related to the traffic congestion, unnecessary time spent searching for parking spots, and, consequently, environmental issues. These systems intend to help drivers that are searching for available parking spaces in a given desired location. This paper presents a cloud-based solution to integrate a Multi-Agent System (MAS) for SP, which enables the modularization, scalability and robustness of such large-scale systems. The MAS abstraction is a suitable approach to represent the dynamic features of a SP, where multiple drivers arrive, request, search, and leave the parking spots. The cloud services enable to scale up the use of a MAS, being an intermediary in the communication between the MAS and the end user, providing a broad architecture that involves database, asynchronous functions activated by events and real-time message exchange. The cloud agent-based system was depl
oyed in the parking of an University campus, where users driving bicycles and cars can request and schedule parking slots that are managed in a distributed manner by the MAS. The obtained results show the user friendly interaction with the system, the scalability of the system in terms of drivers and parking spots, as well as the efficient management of the parking spots by the MAS system.
(More)