Generative AI in Cybersecurity: A Comprehensive Review of LLM Applications and Vulnerabilities

Internet of Things and Cyber-Physical Systems, Год журнала: 2025, Номер unknown

Опубликована: Фев. 1, 2025

Язык: Английский

---

Internet of Things Ontologies for Well-Being, Aging and Health: A Scoping Literature Review

Electronics, Год журнала: 2025, Номер 14(2), С. 394 - 394

Опубликована: Янв. 20, 2025

Internet of Things aims to simplify and automate complicated tasks by using sensors other inputs for collecting huge amounts data, processing them in the cloud on edge networks, allowing decision making toward further interactions via actuators outputs. As connected IoT devices rank billions, semantic interoperability remains one permanent challenges, where ontologies can provide a great contribution. The main goal this paper is analyze state research well-being, aging, health services ontologies. This was achieved analyzing following questions: “Which have been used implement aging services?” “What dominant approach achieve solutions health?’ We conducted scoping literature review papers from 2013 2024 applying PRISMA-ScR meta-analysis methodology with custom-built software tool an exhaustive search through digital libraries: IEEE Xplore, PubMed, MDPI, Elsevier ScienceDirect, Springer Nature Link. By thoroughly 30 studies initial pool more than 80,000 studies, we conclude that increasingly adopt Semantic Web standards integrating heterogeneous data unified models. Emerging approaches, like communication, Large Language Models Edge Intelligence, sustainability-driven analytics, enhance service efficiency promote holistic “One Well-Being, Aging, Health” framework.

Язык: Английский

---

Twin Delayed DDPG (TD3)-Based Edge Server Selection for 5G-Enabled Industrial and C-ITS Applications

IEEE Open Journal of the Communications Society, Год журнала: 2025, Номер 6, С. 3332 - 3343

Опубликована: Янв. 1, 2025

Язык: Английский

---

Urban Road Anomaly Monitoring Using Vision–Language Models for Enhanced Safety Management

Applied Sciences, Год журнала: 2025, Номер 15(5), С. 2517 - 2517

Опубликована: Фев. 26, 2025

Abnormal phenomena on urban roads, including uneven surfaces, garbage, traffic congestion, floods, fallen trees, fires, and accidents, present significant risks to public safety infrastructure, necessitating real-time monitoring early warning systems. This study develops Urban Road Anomaly Visual Large Language Models (URA-VLMs), a generative AI-based framework designed for the of diverse road anomalies. The InternVL was selected as foundational model due its adaptability this purpose. URA-VLMs features dedicated modules anomaly detection, flood depth estimation, level assessment, utilizing multi-step prompting retrieval-augmented generation (RAG) precise adaptive analysis. A comprehensive dataset 3034 annotated images depicting various scenarios developed evaluate models. Experimental results demonstrate system’s effectiveness, achieving an overall detection accuracy 93.20%, outperforming state-of-the-art models such InternVL2.5 ResNet34. By facilitating decision-making, AI approach offers scalable robust solution that contributes smarter, safer environment.

Язык: Английский

---

Domain-Specific Manufacturing Analytics Framework: An Integrated Architecture with Retrieval-Augmented Generation and Ollama-Based Models for Manufacturing Execution Systems Environments

Processes, Год журнала: 2025, Номер 13(3), С. 670 - 670

Опубликована: Фев. 27, 2025

To support data-driven decision-making in a Manufacturing Execution System (MES) environment, system that can quickly and accurately analyze wide range of production, quality, asset, material information must be deployed. However, existing MES data management approaches rely on predefined queries or report templates lack flexibility limit real-time decision support. In this paper, we proposes domain-specific Retrieval-Augmented Generation (RAG) architecture extends LangChain’s capabilities with (MES)-specific components the Ollama-based Local Large Language Model (LLM). The proposed addresses unique requirements including sensor processing, complex manufacturing workflows, knowledge integration. It implements three-layer structure: an application layer using FastAPI for high-performance asynchronous LLM natural language understanding, storage combining MariaDB, Redis, Weaviate efficient management. effectively handles MES-specific challenges such as schema relationships, temporal security concerns without exposing sensitive factory data. This is industry-specific, customized approach focusing problem-solving sites, going beyond simple text-based RAG. considers specificity sources, high-availability requirements, reflection domain compliance quality control regulations, direct interoperability systems. further enhanced through integration various systems, advanced LLM, distributed processing frameworks while maintaining its core focus specialization.

Язык: Английский

---

Generative AI and LLMs for Critical Infrastructure Protection: Evaluation Benchmarks, Agentic AI, Challenges, and Opportunities

Sensors, Год журнала: 2025, Номер 25(6), С. 1666 - 1666

Опубликована: Март 7, 2025

Critical National Infrastructures (CNIs)—including energy grids, water systems, transportation networks, and communication frameworks—are essential to modern society yet face escalating cybersecurity threats. This review paper comprehensively analyzes AI-driven approaches for Infrastructure Protection (CIP). We begin by examining the reliability of CNIs introduce established benchmarks evaluating Large Language Models (LLMs) within contexts. Next, we explore core issues, focusing on trust, privacy, resilience, securability in these vital systems. Building this foundation, assess role Generative AI LLMs enhancing CIP present insights applying Agentic proactive defense mechanisms. Finally, outline future directions guide integration advanced methodologies into protecting critical infrastructures. Our provides a strategic roadmap researchers practitioners committed fortifying national infrastructures against emerging cyber threats through synthesis current challenges, benchmarking strategies, innovative applications.

Язык: Английский

---

Enhancing Bottleneck Analysis in Ship Manufacturing with Knowledge Graphs and Large Language Models

Machines, Год журнала: 2025, Номер 13(3), С. 224 - 224

Опубликована: Март 10, 2025

Ship manufacturing is a critical backbone industry in China, where the nation leads on global scale terms of vessel completions and order volumes. However, high volume orders often imposes substantial processing loads, increases risk equipment failures, exacerbates production bottlenecks. Despite accumulation significant amounts data this field, analyzing bottlenecks remains persistent challenge, primarily due to presence heterogeneous, multi-source lack effective integration mechanisms. The traditional approaches are largely limited bottleneck detection, offering minimal capabilities deep analysis, traceability, interpretability, which crucial for comprehensive resolution. Meanwhile, extensive knowledge underutilized, leading analytical results that overly reliant expert experience lacking interpretability. To address these challenges, research proposes graph-retrieval-based mining method ship manufacturing, employing large language models graph. approach integrates data-driven “turning point” mechanism dynamic detection process graph, consisting subgraphs 5M1E (Man, Machine, Material, Method, Measurement, Environment) specification subgraphs. Furthermore, question-answering chain introduced enhance interaction between LLMs improving retrieval reasoning capabilities. Using practical from Shanghai thin plate line, our demonstrates superior performance compared four existing models, validating its effectiveness throughput analysis. This provides scalable efficient solution complex issues industrial production, contributing enhanced efficiency digital transformation.

Язык: Английский

---

Deploying Large Language Model on Cloud-Edge Architectures: A Case Study for Conversational Historical Characters

Lecture notes on data engineering and communications technologies, Год журнала: 2025, Номер unknown, С. 196 - 205

Опубликована: Янв. 1, 2025

Язык: Английский

---

Environmental AI for Resilient and Sustainable Urban Infrastructures: Rethinking Smart City Transformation

Sustainable civil infrastructures, Год журнала: 2025, Номер unknown, С. 552 - 562

Опубликована: Янв. 1, 2025

Язык: Английский

---

When IoT Meet LLMs: Applications and Challenges

2021 IEEE International Conference on Big Data (Big Data), Год журнала: 2024, Номер unknown, С. 7075 - 7084

Опубликована: Дек. 15, 2024

Язык: Английский

---