Sustainable smart cities require cross-domain, adaptive, and anticipatory intelligence. Current Internet of Things and Artificial Intelligence systems are largely siloed, limiting their ability to support comprehensive urban management. This review addresses this gap by examining the integration of Artificial General Intelligence with the Cyber-Physical-Social-Thinking paradigm. We synthesize research published from January 2015 to September 2025 to provide an overview of how Artificial General Intelligence can serve as the cognitive core, facilitating proactive orchestration across Internet of Things-driven urban layers. Specifically, the paper (1) presents a conceptual analysis of reasoning and meta-cognitive capabilities within the Cyber-Physical-Social-Thinking framework, (2) maps enabling technologies, including digital twins, neuro-symbolic Artificial Intelligence, and the edge-cloud continuum, that operationalize this vision, and (3) reviews applications in mobility, energy management, and climate resilience, highlighting potential gains in efficiency, adaptability, and sustainability. Key technical, ethical, and governance challenges, such as scalability, security, and algorithmic fairness, are discussed. By integrating these insights, the review provides a unified perspective and a roadmap for future research on human-centric, resilient, and sustainable urban intelligent systems.

The integration of the Internet of Everything (IoX) and emerging Artificial General Intelligence (AGI) has given rise to a transformative paradigm aimed at addressing critical bottlenecks across the sensing, network, and application layers in Cyber-Physical-Social-Thinking (CPST) ecosystems. In this survey, we provide a systematic and comprehensive review of pre-AGI and AGI-inspired approaches for IoX, focusing on three key components: sensing-layer data management, network-layer protocol optimization, and application-layer decision-making frameworks. Specifically, this survey explores how pre-AGI and AGI-inspired strategies can mitigate IoX bottlenecks by leveraging adaptive sensor fusion, edge preprocessing, and selective attention mechanisms at the sensing layer. At the network layer, the survey examines solutions to challenges such as protocol heterogeneity and dynamic spectrum management, including approaches based on neuro-symbolic reasoning, active inference, and causal reasoning. Furthermore, the survey investigates AGI-inspired frameworks for managing identity and relationship explosion at the application layer. Key findings suggest that emerging AGI-inspired approaches offer novel solutions to sensing-layer data overload, network-layer protocol heterogeneity, and application-layer identity explosion. These solutions include adaptive sensor fusion, edge preprocessing, and semantic modeling. The survey underscores the importance of cross-layer integration, quantum-enabled communication, and ethical governance frameworks for future AGI-driven IoX systems. Finally, the survey identifies unresolved challenges, including computational requirements, scalability, and real-world validation, and calls for further research to fully realize AGI’s potential in addressing IoX bottlenecks. We believe that AGI-enhanced IoX is emerging as a critical research field at the intersection of interconnected systems and advanced AI.

This review explores the evolution from IoT to the Internet of Everything (IoX) within the cyber-physical-social-thinking (CPST) hyperspace, centering on the emerging challenge of ”relationship explosion.” As interconnected systems grow in scale and complexity, the exponential proliferation of internal (e.g., device coordination, data aggregation) and cross-space (e.g., streaming, translating, adapting) relationships leading to scalability, security, and real-time processing challenges. Through a systematic literature review guided by five research questions, we analyze how this relational explosion intensifies across IoX domains—spanning IoT, IoP, and IoTk—and undermines the efficacy of Artificial Narrow Intelligence (ANI) in managing dynamic, heterogeneous environments. This review proposes that Artificial General Intelligence (AGI) offers a transformative solution, enabling adaptive reasoning, cognitive firewalls, and unified decision-making to navigate complex relationship networks. AGI-driven methodologies enhance system resilience, security, and efficiency in aggregating, moderating, and evolving relationships across CPST spaces. The paper outlines a classification of relationship types, evaluates AGI’s advantages over ANI, and proposes a future research roadmap emphasizing ethical governance, human-AGI collaboration, and sustainable architectures. By framing IoX development around the management of relationship explosion, we provide a roadmap for future research, emphasizing interdisciplinary efforts, ethical governance, and sustainable frameworks to foster intelligent, socially aware IoX ecosystems.

Cyberbullying on social media platforms remains a serious threat to digital well-being, requiring intelligent systems capable of detecting both explicit and subtle, emotionally charged abuse. Sentiment analysis (SA) plays a key role by interpreting emotional tone, polarity, and context, offering more nuanced and timely detection than keyword-based models. Emotions like anger, sarcasm, or veiled hostility often precede cyberbullying, especially during impulsive interactions. SA captures these affective cues, improving sensitivity to implicit abuse and coded language. This study presents the first systematic comparison of sentiment-enhanced transformer models such as ALBERT, DeBERTa, ELECTRA, HateBERT, and DeepSeek-coder-1.3b-base, fine-tuned for cyberbullying detection across Twitter (currently X), IMDB, and Amazon. Models were evaluated on predictive performance (Accuracy, Precision, Recall, F1-score), time and cost efficiency (inference time, memory, CPU/GPU use, and energy). ELECTRA + SA outperformed all models, achieving 91.85% accuracy, precision, and recall, and a 91.84% F1-score. It also excelled in efficiency, with 0.069 seconds inference time, 23.92 MB RAM use, 7.2% CPU/GPU usage, and 0.000075 kWh energy consumption, proving highly generalizable, sentiment-sensitive, and suitable for real-time, resource-aware deployment. These results highlight the importance of sentiment integration, dataset diversity, and computational efficiency in building scalable, real-world cyberbullying detection systems.

The effectiveness of cybersecurity research for SCADA systems depends on access to high-quality network traffic data, yet such data remains scarce due to proprietary restrictions and security concerns. Synthetic data generated by machine learning models, particularly Generative Adversarial Networks (GANs), presents a promising alternative. This study provides a preliminary evaluation of GAN-based approaches for SCADA network traffic synthesis using the IEEE ITACHA DNP3 Smart Grid dataset. We compare a general-purpose GAN (CTGAN) with a network-traffic-specific GAN (NetShare) based on fidelity and statistical consistency. Initial results indicate that CTGAN generates statistically diverse synthetic data, while NetShare suffers from excessive duplication, limiting its applicability. These findings offer an early structured roadmap for selecting and refining generative models for SCADA data synthesis, supporting future research in smart grid security.

Biosignal representation learning (BRL) plays a crucial role in emotion recognition for game users (ERGU). Unsupervised BRL has garnered attention considering the difficulty in obtaining ground truth emotion labels from game users. However, unsupervised BRL in ERGU faces challenges, including overfitting caused by limited data and performance degradation due to unbalanced sample distributions. Faced with the above challenges, we propose a novel method of biosignal contrastive representation learning (BCRL) for ERGU, which not only serves as a unified representation learning approach applicable to various modalities of biosignals but also derives generalized biosignals representations suitable for different downstream tasks. Specifically, we solve the overfitting by introducing perturbations at the embedding layer based on the projected gradient descent (PGD) adversarial attacks and develop the sample balancing strategy (SBS) to mitigate the negative impact of the unbalanced sample on the performance. Further, we have conducted comprehensive validation experiments on the public dataset, yielding the following key observations: first BCRL outperforms all other methods, achieving average accuracies of 76.67%, 71.83%, and 63.58% in 1D-2 C Valence, 1D-2 C Arousal, and 2D-4 C Valence/Arousal, respectively; second, the ablation study shows that both the PGD module (+7.58% in accuracy on average) and the SBS module (+14.60% in accuracy on average) have a positive effect on the performance of different classifications; third, BCRL model exhibits the certain generalization ability across the different games, subjects and classifiers.

Biosignal sequence real-time prediction (BSRP) is essential for predicting the future emotional experience of game users. However, BSRP for game users faces challenges, including poor real-time performance and limited feature fusion dimensions. To address these issues, we proposed a method for BSRP based on the features fusion of local–global and time–frequency domain (LGTF) for game users, which integrates real-time capabilities with multidimensional features fusion. Specifically, LGTF meets real-time requirements and achieves the features fusion of local–global (LG) through multichannel synchronized adaptive convolution. In addition, LGTF implements the features fusion of interband and intraband in the frequency domain and the features fusion of time–frequency (TF) domain by incorporating the self-attention mechanism and Fourier Transform. Furthermore, we conducted comprehensive validation experiments on LGTF using the public dataset. The results indicate that: first, in the comparison study, LGTF outperformed other methods, achieving the lowest average mean squared error (MSE) and mean absolute error values across different prediction lengths of 0.61 and 0.47, respectively. Second, ablation studies revealed that the addition of TF domain feature fusion and LG feature fusion both have the positive effect on the prediction performance, reducing the average MSE by 0.11 and 0.09, respectively. Third, generalization study shows that LGTF exhibits stable performance and generalization across different subjects and shows performance advantages in specific game scenarios. Fourth, time performance analysis suggests LGTF has the real-time performance. Finally, case study demonstrates that LGTF is practical for predicting game users' future emotions and enhancing their emotional experiences.

Cyberbullying significantly impacts mental health by adversely affecting victims’ psychological well-being. It is a prevalent issue on social media platforms, necessitating effective real-time detection systems to identify harmful content. However, current detection systems face challenges related to performance, dataset quality, time efficiency, and computational costs. This study compares existing text classification techniques for cyberbullying detection, evaluating their effectiveness on social media platforms. Large language models such as BERT, RoBERTa, XLNet, DistilBERT, and GPT-2.0 are assessed for their suitability. Results show that BERT achieves optimal performance, with 95% accuracy, precision, recall, and F1 score; a 5% error rate; 0.053 seconds inference time; 35.28 MB RAM usage; 0.4% CPU/GPU utilization; and 0.000263 kWh energy consumption. These findings highlight that while generative AI models are powerful, fine-tuned models often outperform them when adapted to specific datasets and tasks.

Sustainable AI considers both the environmental impact of AI technologies and their role in advancing global sustainability. This study provides a comprehensive analysis of emerging trends in sustainable AI, highlighting energy-efficient algorithms, green data centers, AI-driven resource management, and ethical governance. It emphasizes AI's dual nature, its potential to drive climate action and socio-economic progress, alongside risks such as carbon emissions, e-waste, and digital inequality. The study underscores the need for interdisciplinary collaboration, inclusive policies, and standardized sustainability metrics to ensure responsible AI deployment. By identifying barriers and proposing strategies, it offers guidance to researchers, policymakers, and industry leaders on aligning AI with long-term environmental, ethical, and social objectives.

Autonomous driving systems critically depend on accurately classifying traffic signs, a task that is jeopardized by adversarial attacks. This paper focuses on the relatively unexplored domain of Traffic Sign Classification (TSC) in the context of detecting adversarial attacks. We conduct a rigorous evaluation of six prominent adversarial example detection methods, each representing a distinct detection category and well-regarded within the research community. Our methodology involves adapting these techniques, originally developed for general image classification (IC), to the unique challenges posed by traffic sign images, characterized by their complexity due to factors like varying environmental conditions and a large number of classes. Our study reveals insights into their effectiveness, with the Natural Scene Statistics (NSS) method outperforming others with 83.42%, 86.42%, and 99.96% detection rates; 0.08%, 0.05%, and 0.04%, false positive rates; and 0.02sec, 0.01sec, and 0.02sec detection times for Chinese, Belgium and German traffic sign datasets, respectively. NSS’s superiority is crucial for autonomous vehicles. Our study also sheds light on the often-neglected aspect of detection time in IC, which plays a vital role in ensuring operational efficiency and safety for autonomous vehicles. Our research highlights the need for customized defense strategies tailored to the TSC domain, considering our evaluation’s findings. By identifying promising techniques for detecting adversarial attacks in TSC, we contribute to enhancing the safety and robustness of autonomous driving systems. This study fills a critical knowledge gap, providing valuable insights into understanding and defending against adversarial attacks, specifically in the TSC context.

Intelligent transportation systems (ITS) have made significant advancements in enhancing transportation safety, reliability, and efficiency. However, challenges persist in security, privacy, data management, and integration. Metaverse, an emerging technology enabling immersive and simulated experiences, presents promising solutions to overcome these challenges. By establishing secure communication channels, facilitating virtual simulations for safe testing and training, and enabling centralized data management with real-time analytics, metaverse offers a transformative approach to address these challenges. While metaverse has found extensive applications across industries, its potential in transportation remains largely untapped. This comprehensive review delves into the integration of the metaverse in ITS, exploring key technologies like virtual reality, digital twin, blockchain, and artificial intelligence, and their specific applications in the context of ITS. Real-world case studies, research projects, and initiatives are compiled to showcase the metaverse’s potential for ITS. It also examines the societal, economic, and technological implications of metaverse integration in ITS and highlights the associated integration challenges. Lastly, future research directions are identified to unlock the metaverse’s full potential in enhancing transportation systems.

In recent years, the concept of the Metaverse has attracted considerable attention. This article provides a comprehensive overview of the Metaverse. First, the development status of the Metaverse is presented. We summarize the policies of various countries, companies, and organizations relevant to the Metaverse, as well as statistics on the number of Metaverse-related publications. Characteristics of the Metaverse are identified: 1) multitechnology convergence; 2) sociality; and 3) hyper-spatio-temporality. For the multitechnology convergence of the Metaverse, we divide the technological framework of the Metaverse into five dimensions. For the sociality of the Metaverse, we focus on the Metaverse as a virtual social world. Regarding the characteristic of hyper-spatio-temporality, we introduce the Metaverse as an open, immersive, and interactive 3-D virtual world which can break through the constraints of time and space in the real world. The challenges of the Metaverse are also discussed.

Emotional arousal, an essential dimension of game users’ experience, plays a crucial role in determining whether a game is successful. Game users’ emotion arousal assessment (GUEA) is of great importance. However, GUEA often faces challenges, such as selecting emotion-inducing games, labeling emotional arousal, and improving accuracy. In this study, the scheme for verifying the effectiveness of emotion-induced games is proposed so that the selected games can induce the target emotions. In addition, the personalized arousal label generation method is developed to reduce the errors caused by individual differences among subjects. Furthermore, to improve the accuracy of GUEA, the Breath Rate Variability (BRV) signal is used as a GUEA indicator along with commonly used physiological signals. Comparative experiments on GUEA based on multimodal physiological signals are conducted. The experimental result shows that the accuracy of GUEA is improved by adding the BRV signal, up to 92%.

Modern security practices promote quantitative methods to provide prioritisation insights and support predictive analysis, which is supported by open-source cybersecurity databases such as the Common Vulnerabilities and Exposures (CVE), the National Vulnerability Database (NVD), CERT, and vendor websites. These public repositories provide a way to standardise and share up-to-date vulnerability information, with the purpose to enhance cybersecurity awareness. However, data quality issues of these vulnerability repositories may lead to incorrect prioritisation and misemployment of resources. In this paper, we aim to empirically analyse the data quality impact of vulnerability repositories for actual information technology (IT) and operating technology (OT) systems, especially on data inconsistency. Our case study shows that data inconsistency may misdirect investment of cybersecurity resources. Instead, correlated vulnerability repositories and trustworthiness data verification bring substantial benefits for vulnerability management.

The present paper considers emerging Internet of Things (IoT) applications and proposes a Long Short Term Memory (LSTM) based neural network for predicting the end of the broadcasting period under slotted CSMA (Carrier Sense Multiple Access) based MAC protocol and Energy Harvesting enabled Wireless Networks (EHWNs). The goal is to explore LSTM for minimizing the number of missed nodes and the number of broadcasting time-slots required to reach all the nodes under periodic broadcast operations. The proposed LSTM model predicts the end of the current broadcast period relying on the Root Mean Square Error (RMSE) values generated by its output, which (the RMSE) is used as an indicator for the divergence of the model. As a case study, we enhance our already developed broadcast policy, ADAPCAST by applying the proposed LSTM. This allows to dynamically adjust the end of the broadcast periods, instead of statically fixing it beforehand. An artificial data-set of the historical data is used to feed the proposed LSTM with information about the amounts of incoming, consumed, and effective energy per time-slot, and the radio activity besides the average number of missed nodes per frame. The obtained results prove the efficiency of the proposed LSTM model in terms of minimizing both the number of missed nodes and the number of time-slots required for completing broadcast operations.

Relevant research has shown the potential to understand the game user experience (GUX) more accurately and reliably by measuring the user’s psychophysiological responses. However, the current studies are still very scarce and limited in scope and depth. Besides, the low-detection accuracy and the common use of the professional physiological signal apparatus make it difficult to be applied in practice. This article analyzes the GUX, particularly flow experience, based on users’ physiological responses, including the galvanic skin response (GSR) and heart rate (HR) signals, captured by low-cost wearable devices. Based on the collected data sets regarding two test games and the mixed data set, several classification models were constructed to detect the flow state automatically. Hereinto, two strategies were proposed and applied to improve classification performance. The results demonstrated that the flow experience of game users could be effectively classified from other experiences. The best accuracies of two-way classification and three-way classification under the support of the proposed strategies were over 90% and 80%, respectively. Specifically, the comparison test with the existing results showed that Strategy1 could significantly reduce the negative interference of individual differences in physiological signals and improve the classification accuracy. In addition, the results of the mixed data set identified the potential of a general classification model of flow experience.

Power grids form the central critical infrastructure in all developed economies. Disruptions of power supply can cause major effects on the economy and the livelihood of citizens. At the same time, power grids are being targeted by sophisticated cyber attacks. To counter these threats, we propose a domain-specific language and a repository to represent power grids and related IT components that control the power grid. We apply our tool to a standard example used in the literature to assess its expressiveness.