[IEEE Xplore] Readings on Smart Cities -- [Editorial] Vol. 2, Issue 1, January 2016
Agent-Based Smart Cities
By Rosaldo J. F. Rossetti
The urban smartification phenomenon has yielded profound innovation and cultural changes in society as citizens become aware of and empowered with information and communication technologies (ICT). Given such an ICT-rich environment, even more emphasised with the advent of the Internet of Things (IoT), smart cities have naturally evolved into cyber-physical socio-technical systems. In search of appropriate metaphors to represent this very nature, the present issue of our Readings on Smart Cities introduces three papers to start a discussion on how multi-agent systems (MAS) can contribute to the smartification process of urban settings. Formed up of intelligent autonomous agents with individual and collective goals, and endowed with social abilities, multi-agent systems have been highly applied to tackle problems that are inherently distributed, decentralised, and dynamic, which are very characteristics of smart cities.
As an attempt to meet the European Community’s demand for a scalable, portable and expressive model underlying the conceptualisation of smart cities, Roscia, Longo, and Lazaroiu  propose an approach based on MAS and IoT. Authors consider that, to become smart, cities must be built upon collaborative and interactive subsets, which basically are the different dimensions of an urban ecosystem. They detail their intelligent distributed autonomous smart city (IDASC) model, outlining subsystems, technologies, operating systems and applications to leverage urban smartification. Ponte and colleagues  emphasise on the importance of efficient management of scarce resources in smart cities. Their work is rather on the application perspective of multi-agent systems, as authors devise and implement agents capable of performing complex forecasting methods to improve decision-making in water management. Considering water as a vital resource to urban settings, authors discuss on how to estimate demand to optimally manage resource distribution and use as agents interact and negotiate. In the third article of this issue, multi-agent systems are also approached from an applied point of view. Hurtado, Nguyen, and Kling  idealise power systems on a rather decentralised and horizontal structure allowing their building blocks to interact autonomously so as to ensure energy efficiency and sustainability. Authors propose an agent-based control strategy for the operation of buildings leveraged on the exchange of relevant information so that energy efficiency can be dynamically achieved and improved.
The characteristics of a smart city ecosystem naturally suggest a huge application potential of multi-agent systems, which is already recognised by the scientific community. Also, it is quite intuitive that the multiple dimensions foreseen by virtually all smart city projects constitute smart subsystems on their own, forming up a fabric of interacting self-contained organisations. Such a holonic perspective of smart cities might enormously benefit from MAS-based approaches and technologies, opening up a vast array of cross-fertilisation opportunities in both domains.
IEEE Xplore References
- M. Roscia, M. Longo and G. C. Lazaroiu, "Smart City by multi-agent systems," Renewable Energy Research and Applications (ICRERA), 2013 International Conference on, Madrid, 2013, pp. 371-376.
- B. Ponte, D. de la Fuente, R. Pino and P. Priore, "Multiagent system for intelligent Water Demand Management," New Concepts in Smart Cities: Fostering Public and Private Alliances (SmartMILE), 2013 International Conference on, Gijon, 2013, pp. 1-7.
- L. A. Hurtado, P. H. Nguyen and W. L. Kling, "Agent-based control for building energy management in the smart grid framework," IEEE PES Innovative Smart Grid Technologies, Europe, Istanbul, 2014, pp. 1-6.
Access to articles is complimentary to members of the IEEE Smart Cities Technical Community for a period of 30 days. Please subscribe to receive our selected articles each month via email.