Journal of Information Technology and Sciences

AURA – Advanced Unified Responsive Assistant for Medical Emergencies

2026-09-09T08:18:32+00:00

Modern healthcare systems often struggle with slow disease detection, disconnected patient care, and delayed emergency responses. To address these challenges, this paper presents AURA (Advanced Unified Responsive Assistant for Medical Emergencies), an integrated healthcare platform designed to provide intelligent, accessible, and timely medical assistance. The proposed system combines four major functionalities within a single application: a machine learning based disease prediction module, an AI-powered medical chatbot for answering general healthcare queries, a real-time SOS emergency feature that shares the user’s live location with nearby hospitals, and a medical supply chain tracking module for monitoring the distribution of essential healthcare resources during emergencies and disease outbreaks. Unlike conventional healthcare applications that focus on individual services, AURA unifies these components into a comprehensive platform to improve early diagnosis, emergency response, and healthcare accessibility. This survey reviews recent research related to these technologies, identifies existing limitations, and provides the foundation for developing the proposed integrated healthcare solution.

A Unified TSN–CPS (Time-Sensitive Networking–Cyber-Physical Systems) Framework for Mobile Communication

2026-10-05T10:41:05+00:00

The convergence of operational technology (OT) and mobile telecommunications requires hard deterministic guarantees for untethered physical assets operating within closed-loop feedback systems. Traditional wired Time-Sensitive Networking (TSN) standards (IEEE 802.1) deliver bounded latency and sub-microsecond synchronization but lack support for physical mobility. Conversely, standard wireless networks introduce stochastic jitter and scheduling delays incompatible with hard real-time Cyber-Physical Systems (CPS). This paper presents a Unified TSN–CPS Framework for Mobile Communication that seamlessly integrates IEEE 802.1 TSN wireline standards with 5G-Advanced and emerging 6G wireless architectures. By deploying Device-Side and Network-Side TSN Translators (DS-TT/NW-TT) alongside 5G Time Synchronization Systems (5TSS), the framework abstracts the wireless air interface into a transparent IEEE 802.1Q virtual bridge. The study formulates a cross-layer joint scheduling optimization model that maps wired Gate Control Lists (IEEE 802.1Qbv) to Configured Grant uplink radio resources. Experimental and co-simulation results demonstrate that the proposed framework restricts worst-case latency to under , holds end-to-end jitter within , and maintains clock synchronization accuracy to within . Furthermore, integrating IEEE 802.1CB Frame Replication and Elimination for Reliability (FRER) over dual wireless paths eliminates control loop disruptions during cell handovers, achieving packet delivery reliability. The framework establishes a scalable blueprint for untethered mission-critical applications, including autonomous mobile robots, cooperative vehicle platooning, and industrial tele-robotics.

File Hash Integrity Checker for Digital Forensics and Data Verification

2026-10-15T10:36:15+00:00

File Hash Integrity Checker is a lightweight desktop application designed to verify the integrity and authenticity of digital files through cryptographic hashing. Built using Python and Tkinter, the tool allows users to generate hash values (MD5, SHA-256, and SHA-512) for any selected file and compare them against a known reference hash to detect unauthorized modifications, corruption, or tampering. The application follows a clean separation of concerns, dividing the codebase into a GUI layer (app.py) for user interaction and a logic layer (hash_utils.py) that handles hash computation and comparison using Python's built-in hashlib library. To ensure scalability and reliability, the system reads files in chunks rather than loading them entirely into memory, allowing it to efficiently process files of any size from small text documents to large binary files without performance degradation. Upon computing a file's hash, the application instantly alerts the user whether the generated hash matches a provided reference value, offering a quick and reliable way to confirm file integrity. This makes the tool particularly useful for verifying software downloads, checking for file corruption during transfer, and validating that sensitive files have not been altered. Overall, this project demonstrates practical application of cryptographic principles, GUI development, and efficient file handling in Python, providing a simple yet effective solution for everyday file verification needs. The modular project structure and use of version control (Git/GitHub) also reflect standard software engineering practices suitable for further extension, such as adding drag-and-drop support, batch hashing, or additional hash algorithms.

Performance Evaluation of CNN for Industrial Steel Defect Classification

2026-10-21T10:20:56+00:00

Surface inspection is a critical stage of steel manufacturing, and manual visual inspection is slow, subjective, and hard to scale. This study aims to construct a convolutional neural network (CNN) pipeline that classifies the six types of hot-rolled steel surface defects (crazing, inclusion, patches, pitted surface, rolled-in scale, and scratches) based on the public NEU surface-defect dataset (1,800 grayscale images, 300 images of each type). Three baseline networks, a basic CNN, a ResNet-style network and a MobileNet-style network, were first compared under a fixed 70/15/15 split; they reached 74.1%, 58.1% and 74.8% test accuracy, and the deeper network overfitted badly. To close this gap, the study designed a compact network built from batch-normalised depthwise-separable convolutions with global average pooling (66,198 parameters), trained from scratch with stronger augmentation and a cosine learning-rate schedule. Over three random seeds on the official split, it achieved 97.9 ± 0.2% accuracy (macro-F1 0.979), and stratified five-fold cross-validation on all 1,800 images gave 98.7 ± 0.7% (95% confidence interval 97.8–99.6%, macro-F1 0.987). Class-wise precision, recall, and F1, and the pooled confusion matrix show that inclusion and pitted surface remain the most confused pair. Full hyperparameters and the evaluation protocol are reported so that the experiments can be repeated.

The Evolution of Artificial Intelligence: Emerging Technologies, Applications, and Challenges

2026-11-02T05:28:53+00:00

In this study, the history of artificial intelligence and its technological evolution over time are systematically examined. The study traces the comprehensive trajectory of AI starting from its inception in symbolic logic, rule-based reasoning, and early expert systems to modern data-driven paradigms. Over the decades, AI has undergone multiple fundamental shifts, progressing from manually engineered heuristics to statistical machine learning, and subsequently toward deep representation learning and large-scale foundation models. In the years leading up to 2026, the field has experienced a decisive transition from passive, single-turn generative systems to autonomous agentic architectures capable of multi-step planning, iterative reasoning, and dynamic tool orchestration. Alongside standard dense Transformer models, the study evaluates emerging sub-quadratic architectural alternatives, specifically State Space Models and sparse Mixture-of-Experts, designed to alleviate computational bottlenecks, mitigate context-length limitations, and enhance memory efficiency in long-sequence processing. Furthermore, the deployment of edge computing paired with optimized small language models is investigated as a viable strategy to achieve decentralized intelligence and reduce operational reliance on centralized, power-intensive cloud infrastructure. The study also analyzes cross-domain real-world applications across healthcare, software engineering, and cybersecurity, identifying concrete operational improvements along with domain-specific deployment constraints. Finally, critical sociotechnical challenges are examined in detail, covering algorithmic alignment, risks of model collapse from synthetic data degradation, excessive carbon footprints in large training clusters, and compliance with emerging international governance frameworks such as the EU AI Act. By synthesizing foundational history with contemporary architectural milestones, this paper highlights persistent research gaps and outlines key directions toward scalable, reliable, and embodied AI systems.