Energy-Efficient AI Video Analytics with MPEG-5 LCEVC: GPU-Optimised Results

In the present situation, applications related to multimedia are discovered to be comfortable with the use of video. The number of end consumers who use video continues to rise every day. People are presently searching for videos with better quality among the ones that are currently available there. This results in the launch and dissemination of HD (high definition) videos. Ultra high-definition (UHD) videos are becoming more and more popular as a result of this advancement and need. However, as video communication keeps expanding, there is an upsurge in network traffic because of the limited bandwidth, especially among smart cities. Different advancement codecs have been suggested to deal with the data stream to overcome this hazardous circumstance. However, the fact that modern UHD videos have huge amounts of data makes the available codecs even more complicated. UHD videos can be processed with the latest improvement codec, H.265/High-efficiency video coding (HEVC). Nevertheless, it is impacted by increased power consumption and intricate calculations. Limitations in the codec's functionality confine its use to specific applications, preventing its application in wireless, mobile, or portable settings. Hence, this research concentrates on implementing frame-level quality enhancement through a deep learning network known as FQE-Net. The deep learning convolutional neural network (DLCNN) is specifically crafted to manage films with resolutions up to 16K. Its primary objectives include reducing complexity, minimizing artifacts, enhancing the efficiency of the HEVC codec, and compacting energy consumption. To achieve superior efficiency, it is imperative to replace the DWT transforms within the HEVC codec with a DLCNN model. Additionally, incorporating the Content Block Search Algorithm for Motion Estimation and Compensation, alongside filtering techniques like Sample Adaptive Filter and Deblocking Filter, becomes essential. The simulation results showed that the suggested FQE-Net performed better than the conventional techniques.

Bandwidth-efficient multi-task AI inference with dynamic task importance for the Internet of Things in edge computing

As Smart Cities development grows, deploying advanced technologies, such as the Internet of Things (IoT), Cyber–Physical Systems, and particularly, Artificial Intelligence (AI), becomes imperative for efficiently managing energy resources. These technologies serve to coalesce elements of the energy life cycle. By integrating smart infrastructures, including renewable energy, electric vehicles, and smart grids, AI emerges as a keystone, improving various urban processes. Using the PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) and the Scopus database, this study meticulously reviews the existing literature, focusing on AI technologies in four principal energy domains: generation, transmission, distribution, and consumption. Additionally, this paper shows the technological gaps when AI is implemented in Smart Cities. A total of 122 peer-reviewed articles are analyzed, and the findings indicate that AI technologies have led to remarkable advancements in each domain. For example, AI algorithms have been employed in energy generation to optimize resource allocation and predictive maintenance, especially in renewable energy. The role of AI in anomaly detection and grid stabilization is significant in transmission and distribution. Therefore, the review outlines trends, high-impact articles, and emerging keyword clusters, offering a comprehensive analytical lens through which the multifaceted applications of AI in Smart City energy sectors can be evaluated. The objective is to provide an extensive analytical framework that outlines the AI techniques currently deployed and elucidates their connected implications for sustainable development in urban energy. This synthesis is aimed at policymakers, urban planners, and researchers interested in leveraging the transformative potential of AI to advance the sustainability and efficiency of Smart City initiatives in the energy sector.

As Artificial Intelligence (AI) is used more and more in digital certification systems, it is important to create stable and efficient environments for the use of Large Language Models (LLMs). AI-based chatbots are very helpful for people who are taking online tests at professional certification schools and for people who are giving tests. However, it is still not clear where the best place is to run AI inference workloads because virtualization can use different amounts of resources and cost different amounts. This study aims to identify the optimal deployment environment by assessing Cloud Init, Virtual Machine (VM), and Linux Container (LXC) within the Proxmox Virtual Environment (VE). This environment tested Ollama and FastAPI on the same hardware (4 vCPU, 16 GB RAM, 32 GB SSD, 80 Mbps) and the Phi3:3.8b model. The study also checked the important numbers like CPU and memory usage, disk and network throughput, latency, and response time. The tests showed that LXC had the fastest disk speed (2.45 MB/s) and network speed (3.33 MB/s). VM had the longest response time (15.64 s) and the longest latency (6.89 ms). Cloud Init had mixed results: it made automation easier but less effective. These results show that the best way to use Cloud Init and LXC together for big certification systems is through hybrid orchestration. This is the best way to get a good balance between AI deployment that is flexible and fast. The methodology section provides a clearer description of the experimental process, including benchmark tools (Hey CLI, Sysbench, Prometheus), the number of test repetitions (three sessions per environment), and comparative data analysis methods to ensure result validity. Moreover, the conclusion emphasizes the scientific implications by explaining how Cloud Init’s automation capabilities can be combined with LXC’s performance efficiency to improve AI inference deployments in scalable and institutional environments.

Smart cities are an emerging technology that is receiving new ethical attention due to recent advancements in artificial intelligence. This paper provides an overview of smart city ethics while simultaneously performing novel theorization about the definition of smart cities and the complicated relationship between (smart) cities, ethics, and politics. We respond to these ethical issues by providing an innovative representation of the agent-deed-consequence (ADC) model in symbolic terms through deontic logic. The ADC model operationalizes human moral intuitions underpinning virtue ethics, deontology, and utilitarianism. With the ADC model made symbolically representable, human moral intuitions can be built into the algorithms that govern autonomous vehicles, social robots in healthcare settings, and smart city projects. Once the paper has introduced the ADC model and its symbolic representation through deontic logic, it demonstrates the ADC model’s promise for algorithmic ethical decision-making in four dimensions of smart city ethics, using examples relating to public safety and waste management. We particularly emphasize ADC-enhanced ethical decision-making in (economic and social) sustainability by advancing an understanding of smart cities and human-AI teams (HAIT) as group agents. The ADC model has significant merit in algorithmic ethical decision-making, especially through its elucidation in deontic logic. Algorithmic ethical decision-making, if structured by the ADC model, successfully addresses a significant portion of the perennial questions in smart city ethics, and smart cities built with the ADC model may in fact be a significant step toward resolving important social dilemmas of our time.

The rapid uptake of blockchain and artificial intelligence (AI) technologies like machine learning (ML) in particular has brought about a paradigm change that is elevating the digital ecosystem for smart cities. The development of new technology can be used to create intelligent societies within smart cities in this era of rapid digital communication. Uses of blockchain technology and ML technologies are proliferating, guaranteeing solutions for issues in a multitude of sectors, such as financial services, cryptocurrency, threat intelligence, social and public services, and the Internet of Things. These “smart cities” are constructed using various technologies, including artificial intelligence and blockchain. As a result, the way these technologies are applied in smart cities both present and future will alter not only the character of governance and human interaction but also the way business is done. The research suggests doing an experimental study to find out how blockchain technology and artificial intelligence are affecting the creation of smart cities. It tries to set the scene for queries like how traditional business models are getting ready for this disruption, what obstacles they might encounter, and what effects both technologies might have on the organization's growth. Blockchain technology and artificial intelligence have a lot of potential to help create smart cities. The results of the research will provide corporations with the rationale to start concentrating on these technologies and putting early adoption strategies into place, which will eventually cause them to change their smart city business models advance.

Artificial intelligence (AI) is embedded in a wide variety of Smart City applications and infrastructures, often without the citizens being aware of the nature of their “intelligence”. AI can affect citizens’ lives concretely, and thus, there may be uncertainty, concerns, or even fears related to AI. To build acceptable futures of Smart Cities with AI-enabled functionalities, the Human-Centered AI (HCAI) approach offers a relevant framework for understanding citizen perceptions. However, only a few studies have focused on clarifying the citizen perceptions of AI in the context of smart city research. To address this gap, we conducted a two-phased study. In the pre-study, we explored citizen perceptions and experiences of AI with a short survey (N = 91). Second, scenario-based interviews (N = 7) were utilized to gain in-depth insights of citizen perceptions of AI in the Smart City context. Five central themes were recognized: (1) I don’t like them monitoring me, (2) I want maximum gain for minimum effort, (3) I don’t want AI to mimic people, (4) I’ll avoid using AI if I consider the risk too high, and (5) I don’t need to be concerned about AI. These offer an idea of human-centered requirements worth considering while designing AI applications for future Smart Cities.

Barriers to artificial intelligence adoption in smart cities: A systematic literature review and research agenda

Cross-sectional adaptability, accuracy, and robustness of embedded and edge computing applications have been facilitated by the growing use of artificial intelligence (AI) in signal processing. However, the real implementation of AI-enhanced signal processing pipes on resource-constrained environments is still a major challenge because there are strict constraints on power consumption, silicon area, memory bandwidth, and real-time latency. The paper has offered a proposal of a set of hardware effective VLSI design methodology used to design AI-enhanced signal processing pipelines specifically designed to run on low-power, resource-limited embedded systems. The suggested approach assumes an algorithm-hardware co-design architecture which narrowly wraps traditional signal preprocessing phases with a slim AI inference device as part of a profoundly-pipelined and parallel VLSI design. In order to realize high hardware efficiency, it uses multiple hardware considerations such as fixed-point arithmetic, quantization-aware model design of AI models, dataflow pipelining that is optimized, and buffering strategies that have low memory usage. The entire architecture is derived on an FPGA platform and tested on post-implementation conditions in order to provide realistic performance evaluation. It has been experimentally shown that the proposed architecture provides significant gains in throughput and energy efficiency over traditional non-optimized and non-AI baseline architectures with competitive inference accuracy. These findings affirm the tight design of AI inference through a hardware-optimized signal processing pipeline to highly decrease the latency and resource overhead. All in all, the suggested VLSI architecture offers a relevant scalable solution to making real-time AI-assisted signal processing possible in resource-constrained embedded and edge systems.

With the rapid development of artificial intelligence (AI), AI for smart decision-making is attracting a lot of attention, but research on this topic in smart cities is not yet comprehensive. Thus, the current research aimed to examine the direct and indirect (via technology anxiety) relationships between artificial intelligence (AI), technology anxiety, and smart decision-making (SDM). The research article also examines the moderating role of internal threats of IoT on AI and technology anxiety and the relationship between AI and smart decision-making. 614 cross-sectional data gathered from participants from public and private sectors in Turkey were utilized to investigate the aforementioned relationships. The results indicated that AI had a positive influence on smart decision-making. AI contributes negatively to technology anxiety. Technology anxiety has a negative effect on smart decision-making. Technology anxiety partially mediated the direct effect of AI on smart decision-making. The results revealed that internal threats of IoT moderated the negative relationship between AI and artificial intelligence, such that the negative relationship is further strengthened when internal threats of IoT are high. The results also indicated that internal threats of IoT moderated the positive direct relationship between AI and smart decision-making, such that the positive relationship is weakened when internal threats of IoT are high. The findings present crucial practical implications for government and local authorities in building smart cities.

Connecting the indispensable roles of IoT and artificial intelligence in smart cities: A survey

The field of the Internet of Things (IoT) is dominating various areas of technology. As the number of devices has increased, there is a need for efficient communication with low resource consumption and energy efficiency. Low Power Wide Area Networks (LPWANs) have emerged as a transformative technology for the IoT as they provide long-range communication capabilities with low power consumption. Among the various LPWAN technologies, Long Range Wide Area Networks (LoRaWAN) are widely adopted due to their open standard architecture, which supports secure, bi-directional communication and is particularly effective in outdoor and complex urban environments. This technology is helpful in enabling a variety of IoT applications that require wide coverage and long battery life, such as smart cities, industrial IoT, and environmental monitoring. The integration of Machine Leaning (ML) and Artificial Intelligence (AI) into LoRaWAN operations has further enhanced its capability and particularly optimized resource allocation and energy efficiency. This systematic literature review provides a comprehensive examination of the integration of ML and AI technologies in the optimization of LPWANs, with a specific focus on LoRaWAN. This review follows the PRISMA model and systematically synthesizes current research to highlight how ML and AI enhance operational efficiency, particularly in terms of energy consumption, resource management, and network stability. The SLR aims to review the key methods and techniques that are used in state-of-the-art LoRaWAN to enhance the overall network performance. We identified 25 relevant primary studies. The study provides an analysis of key findings based on research questions on how various LoRaWAN parameters are optimized through advanced ML, DL, and RL techniques to achieve optimized performance.

This article introduces the image segmentation principles and techniques,and general image segmentation techniques and rule-based inference in artificial intelligence are used to recognize intruders.After segmentation and getting the suspected light-area in the image,the rule-based inference for intelligent recognition of the area is techniquely applied.The target tracking process of inmonitoring system is further introduced.The results show that this technique is practical,reliable and in real-time.It can also send the alarm signal and the suspected light-area image to the alarm center through networks.

Artificial Intelligence (AI) is one of the science fields with huge potential to create a cognitive and tech-leaping type of future smart city design/development. However, extant studies lag behind recent applications, potential growth areas, and the challenges associated with AI implementation. This study examines AI’s current role, trend, and future potential impacts in enhancing smart city drivers. The methodology entails conducting a Systematic Literature Review (SLR) of publications from 2022 onwards. The approach involves qualitative deductive coding methods, descriptive statistical analysis, and thematic analysis. The findings revealed the impacts of AI in (i) public services and connectivity, (ii) improving accessibility and efficiency, (iii) quality healthcare, (iv) education, and (v) public safety. Likewise, strategies, such as collaborative ecosystems, digital infrastructure, capacity building, and clear guidelines and ethical framework, were proposed for fostering the integration of AI in potential future smart cities. This research fills a notable gap in the current understanding of AI’s specific contributions to smart cities, offering insights for stakeholders in urban planning, computer science, sociology, economics, environmental science, and smart city initiatives. It serves as a strategic guideline and scholarly research output for enhancing smart city design. It also underscores the potential of AI in creating dynamic, sustainable, and efficient urban environments.

The rapid expansion of artificial intelligence (AI) systems has intensified concerns regarding their energy consumption and carbon footprint, raising questions about whether efficiency-focused strategies under the Green AI paradigm are sufficient to ensure system-level environmental sustainability. This study systematically synthesizes empirical evidence on the energy use and carbon emissions of AI systems across their life cycle and develops a conceptual framework to integrate sustainability constraints into AI deployment. A systematic review was conducted in accordance with PRISMA 2020 guidelines and AMSTAR-2 standards, with searches performed in Web of Science, Pubmed and Scopus up to 19 December 2025. Eligible studies quantitatively assessed energy consumption, carbon footprint, greenhouse-gas emissions, or life-cycle impacts associated with AI systems, including training, inference, hardware, and deployment infrastructures. Ten studies met the inclusion criteria. The results show that AI-related environmental impacts are substantial and highly context-dependent, with inference-phase energy demand often matching or exceeding training-related consumption in large-scale deployments. Life-cycle assessments indicate that hardware-related emissions and electricity mix strongly influence total carbon footprints, while efficiency gains are frequently constrained by system-level feedback. These findings suggest that isolated efficiency improvements are insufficient and that sustainable AI requires coordinated, system-level governance embedding energy and carbon constraints into design and operational decision-making.