Cognitive smart cities: Challenges and trending solutions
A smart city implies the realization of sustainable city growth enabled by technology-based intelligent solutions to give a good quality of life to its citizens. Information and communication technologies play a crucial role as the nerve centre of the smart city for collecting and analysing data from various sources, like mobile, social media, and sensors. Internet of things (IoT) and big data (BD) also play a critical role in the smart city infrastructures, changing the way we analyse patterns and trends in human behaviour. Smart cities generate a huge amount of data, and therefore need many flexible ways to implement data and processing gateways.
Recently, cognitive analytics have attracted the attention of researchers and practitioners worldwide as a technology-based smart solution. It is a novel approach to information discovery and decision making which uses multiple intelligent technologies such as statistical machine learning, deep learning, distributed artificial intelligence, natural language processing and visual pattern recognition to understand data and generate insights. A cognitive smart city refers to the convergence of emerging IoT and smart city technologies to realize cyber-physical social systems, their generated BD from sensing to communication and computing, and artificial intelligence techniques for all aspects of collaborative computing in sensors, actuators and human-machine interfaces. A cognitive city is one that learns and adapts its behaviour based on the past experiences and can sense, understand and respond to changes of a smart environment with many human and robotic elements. In cognitive cities, data flows not only from the citizens to city management systems (e.g., intelligent transportation systems and healthcare centres), but also from citizen to citizen. Citizens act as human sensors, and intelligence-enabled frameworks build a cyber-physical social system. Thus, consistent citizen engagement, ubiquitous data collection and sophisticated analytics are required to produce the best kind of cognitive city.
The implementation of cognitive smart city is highly context-dependent. Initiatives may range from incremental to disruptive and the deployment is shaped by many factors such as governance, economic, technology, social, environmental, legal and ethical issues. As smart cities projects become more pervasive across geography, technology and applications, it is imperative to identify key learnings to foster a deeper understanding of the technology evolution landscape and provide tangible benefits to smart city planners and key decision-makers. Viable intersection between technology solutions and digital urbanization design principles (people-centred and inclusive infrastructure, resilience and sustainability, interoperability and flexibility, managing risks and ensuring safety) need to be evaluated in order to provide balanced and replicable solutions.
In the research community, several works propose cognitive solutions that fit the needs of sustainable urban development. Academic literature, government consultation documents and policy papers articulate numerous challenges and research directions for incorporating cognition to realize new smart city services. Both qualitative and quantitative studies carefully consider BD analytics, semantic derivation and knowledge discovery, intelligent decision-making and on-demand service provision for a large number of smart city applications.
This special issue aims to stimulate discussion on the design, use and evaluation of self-correction and human cognition for continuous learning as key knowledge discovery drivers within socially connected urban ecosystems. The issue is focused on articles describing cognitive models for cyber–physical social systems with urban BD to leverage deeper insights from the vast amount of generated data with near real-time intelligence.
We received a very good response to our special issue call for articles. During the review process, each article was assigned to and reviewed by at least three experts in the field. After a rigorous multi-round review process, we were able to accept six excellent articles covering various topics related to cognitive smart cities. In the following, we will introduce those articles and highlight their main contributions.
The paper entitled, ‘A taxonomy of energy optimization techniques for smart cities: architecture and future directions’ discussed the need for devising more solutions for efficiently handling energy utilization associated challenges in smart cities. The paper presented a comprehensive survey on the energy optimization techniques in various systems, including the optimization techniques in block chain-based systems. Further, the paper presented a taxonomy that classifies energy optimization techniques and proposed an energy-efficient consensus mechanism, proof-of-high-performance optimization for high-performance computing-based ecosystems.
Authors in the paper, ‘DFT: A deep feature-based semi-supervised collaborative training for vehicle recognition in smart cities’ proposed a deep feature-based training (DFT) method for vehicle recognition in smart cities. DFT is also a semi-supervised collaborative training method on the basis of two base learners. DFT adjusts data pre-processing and training process, optimizes the constructing a disagreement encoding network, and expands on the recognition disagreement of pseudo-labelled samples-based training sets. Compared with the typical collaborative training methods, DFT greatly accelerates the model's training process by reducing the convergence time, and improves the efficiency of vehicle recognition, while remaining the recognition accuracy unchanged.
In the paper, ‘Sensor data fusion for the industrial artificial intelligence of things’ the authors discussed a new framework for addressing the different challenges of the artificial intelligence of things (AIoT) applications. The proposed framework is an intelligent combination of multi-agent systems, knowledge graphs and deep learning. Deep learning architectures are used to create models from different sensor-based data. Multi-agent systems can be used for simulating the collective behaviours of the smart sensors using IoT settings. The communication among different agents is realized by integrating knowledge graphs. Different optimizers based on constraint satisfaction as well as evolutionary computation are also investigated in the paper.
The paper entitled ‘A formal method for privacy-preservation in cognitive smart cities’ presented a discussion on a technique for privacy-preservation in smart cities based on pseudonymization, clustering, anonymization and differential privacy methods. The modified clustering algorithm selects the initial cluster-based on the concept of dissimilarity between the data sequences. The paper also assessed the functional correctness and preformation of the proposed model for privacy-preservation in smart cities.
In the paper, ‘Soft computing for abuse detection using cyber-physical and social BD in cognitive smart cities’ authors discuss the use of soft computing techniques for abuse detection in the complex cyber–physical–social BD systems in cognitive smart cities. The objective of the paper is to define and identify the diverse concept of abuse and systematize techniques for automatic abuse detection for cyber abuse detection on social media and real-time abuse detection using social IoT. The cyber abuse studies on social media platforms have further been categorized as cyber-hate and cyberbullying whereas the real-time abuse includes studies of cyber–physical social systems. As in a cognitive smart city, citizens expect more from their urban environments with minimal intervention, this study helps us to establish the need to capture situational context and awareness in real-time and foster the need to develop a proactive as well as reactive safety mechanism to mitigate the risks of online abuse.
The paper entitled ‘A lightweight intelligent intrusion detection system for industrial IoT using deep learning algorithms’ discusses the techniques for improving intelligent decision-making actions in the industrial IoT (IIoT) network in a sustainable city. Main cybersecurity attacks are predicted by applying a deep learning model in the paper. The various security and integrity features such as Denial of Service (DoS), malevolent operation, data type probing, spying, scanning, intrusion detection, brute force, web attacks and wrong setup are analysed and detected by a novel sparse evolutionary training mechanism-based prediction model.
We would like to express our sincere thanks to all the authors for submitting their papers and to the reviewers for their valuable comments and suggestions that significantly enhanced the quality of the articles. We are also grateful to Editor-in-Chief, Prof. Jon G. Hall and Special Issues & Reviews Editor Prof. Lucia Rapanotti for their great support throughout the whole review and publication process of this special issue, and, of course, all the editorial staff. We hope that this special issue will serve as a useful reference for researchers, scientists, engineers and academics in the field of cognitive smart cities' design and development.
