Computing Reviews

Artificial intelligence, machine learning, and optimization tools for smart cities: designing for sustainability
Pardalos P., Th. Rassia S., Tsokas A., Springer International Publishing,Cham, Switzerland,2022. 243 pp.Type:Book
Date Reviewed: 05/22/23

The book’s 11 chapters deal with digital “vitalism” in cities, unbuildable cities, smart cities and identities, services and data representation of information and communication technology (ICT), strategies for passers-by, electric vehicle sharing (EVS), architectural proposals for fog/edge computing in the Internet of Things (IoT), urban big data such as real estate markets, and social-media-based smart disaster responses. The chapters are useful for looking at the urban fabric for smart cities, including resiliency, responsiveness, elasticity, interconnection, and asynchronicity between components.

Li et al.’s chapter, “A Pedestrian-Level Strategy to Minimize Outdoor Sunlight Exposure,” is a reference to reactive systems such as autopoietic systems--first explored by the German scientist von Bertalanffy in the 1940s, and then strongly reformulated by the Chilean biologists Maturana and Varela in the 1970s--and their ability to pose and then solve themselves by means of the same problematic element for answering new questions in cities, such as reducing potential exposure to sunlight. It can, in fact, be obtained by an algorithmic calculation of the distance of the path of the sun, the destination pairs, and the projection of its position on hemispherical images. This book therefore immerses readers in many fundamental changes that conceptually model the urban fabric through topics related to artificial intelligence (AI) and machine learning (ML).

Ignoring the dependencies, promises, challenges, and algorithmic scenarios of IoT networks in cities, we are entitled, as agents in relation to an environment, to try to decode a sequence of “anticipatory systems,” in the words of the 1950s American scientist Robert Rosen [1], which are believed to have also been expressed over the centuries by such geniuses as Goethe and Kandinsky, as well as Althusser, Bachelard, and Lamarck, to name but a few, as in Moraitis’ chapter “Digital ‘Vitalism’ and its ‘Epistemic’ Predecessors.” An AI outset of a smart city seems to prevail that is no longer tied to a merely mechanized world, made up mostly of tools and weapons, good and bad, buildable and destructible, but rather to a metaphorical self-simulation of what we had been in the past through computational optimization as well as parametric intelligent design. An array of techno-scientific “assets” in cities can certainly stimulate morphogenetic change, and even restore temporary battles in imaginary landscapes that allow us to portray an auspicious representation of our intelligence as we can now reverse evil for good because we identified the key, that is, we are able to intercept the correct flow of models.

In Rassia’s chapter, “Unbuildable Cities,” we find that the ability of ICT to dynamically generate self-embedded boundaries within the organic structures of living systems has limitations, that is, the so-called “ambient connectivity” [2] where humans input, machines create, and the collective efforts of nature cooperatively respond to problem solving through autopoietic operating system (AOS) outputs. We cannot expect to detain universal city models capable of transferring at least some of the well-analyzed AI/ML imagery to urban relics because of man-made and natural calamities. We can approach the problems from a holistic point of view; we can observe from the heights and watch, step by step, how the messages of our city, among the components of the urban fabric, will run executions, address problems and latencies, and tolerate breakdowns. This also suggests that permissions to learn in smart cities are available to everyone without discrimination, but one should thus be informed of the game.

Each ICT intelligence for its part encompasses value chains, service providers, community infrastructure, and government domains as interconnected and open source. Thus, AI components cannot decide peace versus wars, but can calculate the speed, accuracy, and generalizability of responses to problems as they have the ability to interface both publicly and privately with agents to override a secure-by-design architecture that approves inputs through basic consensus. This mechanism, highlighted by Tsoniotis in “Smart Cities as Identities,” is embodied in promising ICT codes that can reinvent identities, creating problem subclasses containing personal data sharing based on voluntary consent. “A Cross-Domain Landscape of ICT Services in Smart Cities,” by Buhnova et al., clearly reinforces how smart cities are interconnected and co-create complex interdependencies that can influence a step change in the development of the urban fabric.

The book is very technical in some parts, for example, “Planning and Management of Charging Facilities for Electric Vehicle Sharing” by He et al. However, the authors’ Bayesian time-frequency adaptive analysis of multivariate time series (MTS) does not remain effortless, like a trompe-l’oeil in the facade of buildings, but on the contrary generates massive reasoning on nonlinear programming (NLP) languages and the relative deployment of city-wide apps and services for the widest range of businesses and people.


Louie, A. H.; , Robert Rosen’s anticipatory systems. Foresight 12, 3(2010), 18–29.


Dufour, P. Control engineering in drying technology: review and trends. Drying Technology 24, 7(2006), 889–904.

Reviewer:  Romina Fucà Review #: CR147592

Reproduction in whole or in part without permission is prohibited.   Copyright 2023™
Terms of Use
| Privacy Policy