IntroductionSoil water management is one of the significant domains in smart irrigation. This paper proposes and implements an IoT and edge computing-based soil water management system for sustainable irrigation using ensemble learning.MethodsAn ultralow power sensor node is used to collect real-time soil moisture and soil temperature data. The edge device performs raw data processing and then sends the pre-processed data to the cloud server after adding Gaussian noise to it for data privacy protection. Ensemble learning models are used for analyzing the data inside the cloud.Results and DiscussionThe results are compared to single classifier-based machine learning models and deep learning models. The results demonstrate that the ensemble learning models outperform the single classifier-based machine learning models in terms of prediction accuracy, and the deep learning models in terms of latency and energy consumption. We also perform soil temperature and soil moisture prediction using ensemble learning to deal with the scenario, where either the soil moisture or soil temperature sensor is available, and the results show that the ensemble learning models outperform the baselines. We also analyze the proposed framework’s performance for other datasets to further demonstrate its efficacy. The results show that the proposed irrigation decision making framework has ≥99% prediction accuracy for our own dataset and publicly available datasets. Finally, an Android App is designed using which the user can access the data remotely, and can control the irrigation decision making process based on the results.
IntroductionRecent advances in artificial intelligence (AI) have positioned large language models (LLMs) as potentially transformative tools for education. However, their adoption in primary schools raises significant concerns regarding student data privacy, institutional governance, and the affordability of AI infrastructure.MethodsThis paper presents a conceptual analysis of locally hosted small language models (SLMs) as an alternative to cloud-dependent AI systems in primary education. The study examines developmental, institutional, and regulatory factors relevant to primary school settings and proposes a school-based architecture integrating low-cost hardware and open-source software frameworks operating entirely within local school networks.ResultsThe proposed architecture demonstrates how locally deployed SLMs in primary schools can support educational AI applications while reducing dependence on external cloud providers. By maintaining data processing and storage within school networks, the framework may strengthen data sovereignty, improve institutional control over AI operations, and enhance accessibility for resource-constrained schools through lower infrastructure costs.DiscussionThe findings suggest that, within primary education, institutional controllability, privacy protection, and equitable access may be more valuable objectives than maximising model capability. This conceptual framework contributes to the emerging literature on educational AI governance by outlining a sociotechnical pathway that prioritises child data protection, institutional autonomy, and practical deployment feasibility while supporting the responsible integration of AI into educational environments.
IntroductionDigital twin technology is emerging as a key enabler of cyber-physical systems, intelligent networks, and data-driven automation. However, research on digital twins is often framed from mechanical or electrical perspectives rather than from computer engineering. Although several comprehensive surveys have addressed enabling technologies, security threats, industrial IoT applications, and business innovation perspectives, none have systematically examined how core computer engineering concerns, including embedded platforms, hardware-software co-design, real-time scheduling, edge-cloud orchestration, and on-device AI, collectively shape the design, deployment, and evaluation of digital twin systems.MethodsTo address this gap, this paper presents a systematic literature review of 45 peer-reviewed publications that explicitly engage with digital twins in computer engineering contexts. The review was conducted using the PRISMA 2020 framework, and inductive thematic analysis was used to identify and synthesize major research themes across the selected studies.ResultsThe analysis revealed six major themes: (i) computer engineering foundations of digital twin systems, including embedded and real-time platforms, hardware-software co-design, and cyber-physical architectures; (ii) networked and distributed infrastructure for digital twins, including IoT, 5G/6G, and edge-cloud computing; (iii) computer engineering for intelligent digital twins, including on-device AI, learning-enabled control, and integration of machine learning with twin models; (iv) security, privacy, and dependability in digital twin-enabled cyber-physical systems; (v) modeling, simulation, and testing of digital twin systems; and (vi) application domains and use cases. The review further shows that networked and AI-enabled digital twins are growing rapidly but remain largely at the prototype stage. Few studies rigorously evaluate digital twin implementations on constrained computing platforms, while interoperability, standardized validation, and safe adaptivity remain persistent challenges.Discussion and ConclusionUnlike prior surveys that catalog digital twin technologies by domain or threat model, this review synthesizes findings through a computer engineering lens. The results highlight the need for hardware-aware, security-integrated, and systematically validated approaches to digital twin design and deployment. The paper concludes by outlining a research agenda to advance the development of robust, interoperable, and scalable digital twin systems within computer engineering.
The convergence of artificial intelligence, sensing technologies, and networked systems is transforming healthcare monitoring beyond traditional clinical settings. Digital twins offer a powerful paradigm for personalized health modeling, while advances in Artificial Intelligence of Things (AIoT) enable intelligent sensing and analysis within everyday environments. However, existing healthcare systems typically treat these domains separately, either employing cloud-centric digital twins or AIoT platforms based on continuous data streaming. To the best of our knowledge, no prior work explicitly integrates digital twin modeling with edge-intelligent AIoT architectures for healthcare, leaving a critical gap between these complementary research areas. Such cloud-centric designs lead to high power consumption, excessive communication bandwidth, privacy risks, and poor scalability for long-term, population-level deployment. This paper presents an edge-intelligent, event-driven AIoT digital twin architecture for scalable and privacy-aware healthcare monitoring. The proposed system performs feature extraction, inference, and event detection locally on resource-constrained AIoT devices, transmitting only semantic health state updates and periodic summaries to maintain digital twins. We evaluate the architecture through computational simulations of a fall and instability monitoring scenario. Results show that the event-driven edge design reduces per-patient communication achieving orders-of-magnitude lower device-side energy consumption and bandwidth usage compared to continuous cloud- and edge-centric baselines. Sensitivity analysis demonstrates robustness across event frequencies, window sizes, and feature extraction costs, while population-scale analysis confirms that system costs scale with event frequency rather than sensor sampling rate. These findings indicate that event-driven edge intelligence is a foundational architectural principle for digital twin healthcare systems, enabling scalable, energy-efficient, and privacy-aware monitoring suitable for long-term real-world deployment.
The growing deployment of IoT devices necessitates reliable sensor data transfer under diverse network conditions. This study introduces an adaptive network switching framework to enhance data dependability across GSM, Wi-Fi, and LoRa networks. By leveraging delay tolerant networking (DTN) principles and real-time performance metrics, the algorithm dynamically selects the optimal channel for transmitting sensor data—text, audio, image, and video—across urban, suburban, and rural settings. Simulations demonstrate an average 33% improvement in throughput, 24% reduction in latency, and 45% decrease in packet loss when using the adaptive framework compared to standalone networks. The adaptability score averaged 0.7 in rural scenarios, with peak performance scores reaching 1,000 for video data at night. A beehive monitoring case study validates these results in real-world conditions. This work contributes a robust, adaptable solution for sensor data optimization in IoT applications.
The widespread adoption of Medical Internet of Things (MIoT) devices, particularly portable electrocardiogram (ECG) monitors, has accelerated since the COVID-19 pandemic, revolutionizing remote patient monitoring and healthcare delivery. However, this rapid integration has introduced significant cybersecurity challenges, especially in securing communication within the MIoT ecosystem. To address these concerns, this study presents a systematic security analysis of three popular portable ECG devices: the Beurer BM 95, KardiaMobile 6L, and OMRON Complete. The investigation begins with a structured literature review to develop a catalog of threats and a threat model specific to the devices’ ecosystem. Guided by this threat model, controlled experiments were conducted to perform penetration testing and security assessments. Our findings reveal multiple security weaknesses and vulnerabilities in the Bluetooth Low Energy (BLE) implementations on these devices, exposing them to potential exploitation and attacks. Additionally, simulated attacks on paired smartphones enabled the recovery of sensitive user and patient data, highlighting further risks within the ecosystem. By uncovering these vulnerabilities, this research highlights the urgent need for stronger security measures in MIoT devices. Addressing these issues proactively is essential to enhance device resilience and protect against emerging threats in connected healthcare environments.
The Autonomous Internet of Things (A-IoT) represents a major advancement in interconnected systems, enabling self-governing smart devices to operate collaboratively across domains such as smart cities, industrial automation, healthcare, and autonomous vehicles. However, the complexity, scale, and heterogeneity of A-IoT environments introduce severe cybersecurity challenges, including expanded attack surfaces, real-time data processing demands, sophisticated adversarial threats, and privacy risks. Traditional security measures are not always adequate to address these emerging threats, and this is why intelligent adaptive defence systems are required. This narrative review offers an extensive and systematic presentation of AI-based cybersecurity strategies that are specific to the peculiarities of A-IoT ecosystems. It examines fundamental methods, including machine learning, deep learning, federated learning, and swarm intelligence, as well as the latest paradigms, such as explainable AI, generative adversarial networks, and digital twins. The approaches are discussed within the scope of the most important security tasks, such as intrusion detection, anomaly detection, malware analysis, secure authentication, and autonomous threat response. The review also locates crucial issues related to data quality, model interpretability, adversarial vulnerabilities and ethical limitations of the application of AI in security-critical applications. Moreover, it describes future research directions using hybrid AI-blockchain frameworks, self-healing autonomous agents, and trust-aware AI systems.
IntroductionThe Internet of Things (IoT) plays a crucial role in realizing the vision of pervasivemcomputing. The variety of devices and services in the IoT often makes manual integration complex and tedious. To mitigate this challenge, self-organization has emerged as a promising approach. Furthermore, the concept of selfexplainability has been proposed to clarify user interactions with dynamically interconnected smart objects.MethodsThis paper explores related research in the domains of (a) service composition, (b) self-explainability, and (c) application management within smart environments. To address existing limitations, we propose a novel application management concept. Therefore, we extended a framework that utilizes self-explaining applications to generate user instructions dynamically that explain the control of these. A user study was conducted to assess the effectiveness and preferred options of modalities of these instructions.ResultsThe findings indicate that the approach successfully tackles the research question concerning the integration of self-explaining applications into smart environments.
ContextThe OPME (Órteses, Próteses e Materias Especiais or Orthoses, Prosthetics and Special Materials) Brazilian sector presents a wide variety of products and technologies, involving both multinational and local companies in healthcare. Despite technological advances, many services and information systems, especially in the public sphere, still use unstructured natural language descriptions of products, services or events, making their classification and analysis difficult. However, for efficient audits, it is necessary to classify and totalize invoices issued for product purchases automatically. In this way, the standardization lacking regarding nomenclature in the OPME marketing not only makes it difficult to compare products, whether for price standardization or standardization of use but also opens up space for possible acts of corruption.ObjectiveTo mitigate the problem of ineffective standardization and coding, develop and assess the effectiveness and efficiency of an OPME classifier, in the context of electronic invoice descriptions, from the point of view of auditors, healthcare professionals, and data scientists.MethodControlled Experiment, to evaluate scientifically mapped Artificial Intelligence (AI) algorithms and compare accuracy measures, F1-Score, sensitivity, precision, average training time, and classification.ResultsWith an accuracy of 99%, the Linear Support Vector algorithm stood out among the others in terms of accuracy, while Naïve Bayes in terms of efficiency, had the fastest average training time.ConclusionThe results showed that it is possible to identify and classify OPMEs in invoices automatically. This allows for a more precise and effective analysis of signs such as anomalously high prices and quantities of OPMEs purchased per inhabitant, which are analyzed by the Audit of Brazil’s Unified Health System (AudSUS), Ministry of Health -Brazil, for identification of potential irregularities and contribution to transparency and efficiency in the management of health resources.
The scarcity of marine life caused by shallow waters and pollution has pushed fishermen to venture up to 10 km offshore, where clearer waters offer richer fish resources, but at much higher fuel costs. This situation has caused increased unemployment, so many fishermen have changed professions. This is exacerbated by seasonal variations and resource limitations. This research introduces a fish sensor network (FSN), which is designed to equip floating fish houses (FADs) with network-connected fish sensors. The proposed network allows fishermen to find locations with high fish populations, thereby reducing fuel costs and increasing fishing efficiency. This article presents preliminary findings, identifies potential challenges including natural factors such as wind, waves, currents, corrosion, and radio propagation, as well as man-made obstacles such as traffic density and physical measurements at the research site in Belawan, Indonesia. Additionally, this paper briefly discusses the energy availability that poses further challenges.
As the number of electronic control units (ECUs) in vehicles continues to grow, exchanging physical components between original equipment manufacturers (OEMs) and ECU suppliers presents logistical challenges, impeding the pace of development and leading to accumulation of costs. In this work, we introduce a conceptual framework that enables remote testing of geographically dispersed ECUs over the Internet. Pertinent to the Internet of Things (IoT) sphere, where interconnection of distributed components is hampered by the inherent challenges of latency, we propose a hybrid synchronous methodology that combines asynchronous test management with time synchronization mechanisms to mitigate the delay impact within the distributed environment. Additionally, we discuss challenges and prospects associated with this approach.
Due to an increase in the number of attacks on Industrial Control Systems (ICS), the security of these systems is now of paramount concern. Many solutions have been proposed to defend such systems, particularly those adopting Machine Learning (ML). The goal of this study is to increase the likelihood of the solution being deployed into a real environment. As part of that, multiple interviews were held with industry experts who have been embedded within ICS cyber-security for decades. The findings revealed that the current security solutions for ICS lack the sophistication required to be adopted due to flawed assumptions made about the end-user. Therefore, this work provides personas of each end-user group within ICS that need to be taken into consideration when designing a security solution. In addition, wireframes are provided showing what a desired solution could look like. By sharing these findings, it is hoped to inform those working within this space and increase the likelihood of their solutions being adopted within a real environment. Furthermore, the expert panel requested a number of features that do not currently exist within the ICS cyber-security space, therefore, by sharing these with the wider community, it is hoped that the field will move closer towards providing solutions containing these features.
This work presents a threat modelling approach to represent changes to the attack paths through an Internet of Things (IoT) environment when the environment changes dynamically, that is, when new devices are added or removed from the system or when whole sub-systems join or leave. The proposed approach investigates the propagation of threats using attack graphs, a popular attack modelling method. However, traditional attack-graph approaches have been applied in static environments that do not continuously change, such as enterprise networks, leading to static and usually very large attack graphs. In contrast, IoT environments are often characterised by dynamic change and interconnections; different topologies for different systems may interconnect with each other dynamically and outside the operator’s control. Such new interconnections lead to changes in the reachability amongst devices according to which their corresponding attack graphs change. This requires dynamic topology and attack graphs for threat and risk analysis. This article introduces an example scenario based on healthcare systems to motivate the work and illustrate the proposed approach. The proposed approach is implemented using a graph database management tool (GDBM), Neo4j, which is a popular tool for mapping, visualising, and querying the graphs of highly connected data. It is efficient in providing a rapid threat modelling mechanism, making it suitable for capturing security changes in the dynamic IoT environment. Our results show that our developed threat modelling approach copes with dynamic system changes that may occur in IoT environments and enables identifying attack paths, whilst allowing for system dynamics. The developed dynamic topology and attack graphs can cope with the changes in the IoT environment efficiently and rapidly by maintaining their associated graphs.
This paper presents a case application of a socio-technical framework for Internet of Things (IoT) research and development in the United Kingdom. Applying a socio-technical system approach to IoT, this paper seeks to provide a clear understanding of the interplay between technical and non-technical aspects of IoT research and development. It describes the socio-technical requirements for IoT design and development and provides the current snapshot of research in the United Kingdom to meet these requirements. Finally, the paper provides useful information on how to conceptualize IoT research within human-centered contexts and a useful guide for centre design and evaluation to those developing new research centres or seeking to reinvigorate existing ones.
The security, privacy, and safety issues around Female-oriented technologies (FemTech) and data can lead to differential harms. These complex risks and harms are enabled by many factors including inadequate regulations, the non-compliant practices of the industry, and the lack of research and guidelines for cyber-secure, privacy-preserving, and safe products. In this paper, we review the existing regulations related to FemTech in the United Kingdom, EU, and Switzerland and identify the gaps. We run experiments on a range of FemTech devices and apps and identify several exploitative practices. We advocate for the policymakers to explicitly acknowledge and accommodate the risks of these technologies in the relevant regulations.
In this work, we introduce the vision of a “Right to Improve”: a hypothetical future law which should entitle consumers to modify and extend Internet of Things devices during the productive lifetime. Current European Union legislation as well as voluntary manufacturer interoperability initiatives fail to address user desires for adaptability, augmentability, and open-ended repurpoposing of Internet of Things (IoT) devices. We therefore argue in this paper for a Right to Improve that aims to fill the gaps left by today’s laws, conforms to consumer demands, and is powerful enough (by its statutory nature) to cause actual change. Our contribution is twofold. First, we summarize the relevant legislation (Ecodesign Directive, Sale of Goods Directive, and Right to Repair). We review the academic literature, and highlight technical and motivational factors that support a Right to Improve. Second, we suggest and discuss a number of open problems in need of consideration by academics, practitioners, companies, governing bodies, and the general public. A future formulation of the Right to Improve should take into account and balance the various contrasting views for efficacy on different levels, e.g., environmental benefit vs. economic risk. Change enabled by a Right to Improve may greatly enhance the usefulness of devices during their use phase, empowers consumers to create and compose devices and services to their own taste and requirements, makes devices more economically and ecologically sustainable by extending their useful lifetime, and creates new business opportunities.
Security in the Internet of Things (IoT) remains a predominant area of concern. This survey updates the state of the art covered in previous surveys and focuses on defending against threats rather than on the threats alone. This area is less extensively covered by other surveys and warrants particular attention. A life-cycle approach is adopted, articulated to form a "defence in depth" strategy against malicious actors compromising an IoT network laterally within it and from it. This study highlights the challenges of each mitigation step, emphasises novel perspectives, and reconnects the discussed mitigation steps to the ground principles they seek to implement.
Particulate matter (PM) is considered the primary contributor to air pollution and has severe implications for general health. PM concentration has high spatial variability and thus needs to be monitored locally. Traditional PM monitoring setups are bulky, expensive, and cannot be scaled for dense deployments. This paper argues for a densely deployed network of IoT-enabled PM monitoring devices using low-cost sensors, specifically focusing on PM10 and PM2.5, the most health-impacting particulates. In this work, 49 devices were deployed in a region of the Indian metropolitan city of Hyderabad, of which 43 devices were developed as part of this work, and six devices were taken off the shelf. The low-cost sensors were calibrated for seasonal variations using a precise reference sensor and were particularly adjusted to accurately measure PM10 and PM2.5 levels. A thorough analysis of data collected for 7 months has been presented to establish the need for dense deployment of PM monitoring devices. Different analyses such as mean, variance, spatial interpolation, and correlation have been employed to generate interesting insights about temporal and seasonal variations of PM10 and PM2.5. In addition, event-driven spatio-temporal analysis is done for PM2.5 and PM10 values to understand the impact of the bursting of firecrackers on the evening of the Diwali festival. A web-based dashboard is designed for real-time data visualization.
In today’s fast-paced world, efficient and reliable public transportation systems are crucial for optimising time and reducing carbon dioxide emissions. However, developing countries face numerous challenges in their public transportation networks, including infrequent services, delays, inaccurate and unreliable arrival times, long waiting time, and limited real-time information available to the users. GPS-based systems have been widely used for fleet management, but they can be a significant infrastructure investment for smaller operators in developing countries. The accuracy of the GPS location can be easily affected by the weather condition and GPS signals are susceptible to spoofing attacks. When the GPS device is faulty, the entire location traces will be unavailable. This paper proposes the use of Internet-of-Things (IoT)-enabled Bluetooth Low Energy (BLE) systems as an alternative approach to fleet tracking for public bus service. The proposed approach offers simplicity and easy implementation for bus operators by deploying BLE proximity beacons on buses to track their journeys, with detection devices using Raspberry Pi (RPi) Zero strategically placed at terminals and selected stops. When the bus approaches and stops at the bus stops, the BLE advertisements emitted by the proximity beacons can be reliably detected by the RPi Zero. Experiment results show that the BLE signals can be detected up to 20 m in range when the RPi Zero is placed inside a metal enclosure. The location of the bus is then sent to the cloud to estimate the arrival times. A field trial of the proposed IoT-based BLE proximity sensing system involving two public bus services in southern Malaysian cities, namely, Johor Bahru , Iskandar Puteri and Kulai is presented. Based on the data collected, a bus arrival time estimation algorithm is designed. Our analysis shows that there was a 5–10 min reduction in journey time on public holidays as compared to a normal day. Overall, the paper emphasises the importance of addressing public transportation challenges. It also describes the challenges, experience, and mitigation drawn from the deployment of this real-world use case, demonstrating the feasibility and reliability of IoT-based proximity sensing as an alternative approach to tracking public bus services.
