Asmau Abubakar Abdulmalik

The rapid growth of multimodal wearable devices has enabled continuous monitoring of physiological and behavioral patterns for home-based health applications. However, centralized data processing raises serious privacy concerns and limits real-time, interpretable insights. This study proposes PEX-FEL, a privacy-preserving and explainable federated edge learning framework for stress detection and activity recognition in decentralized environments. The framework combines federated learning with differential privacy (ε ≤ 1.0) and secure aggregation to protect user data. Low-Rank Adaptation (LoRA) is applied for efficient local training on resource-constrained edge devices, while SHAP is used to provide interpretable, user-centric explanations of predictions. Experiments were conducted using WESAD, PPG-DaLiA, and SWELL datasets under non-IID conditions to simulate real-world heterogeneity. Results show that a hybrid CNN-LSTM model achieved 0.85 accuracy, 0.85 F1-score, and 0.90 AUC-ROC, outperforming centralized approaches by 8–10%. The framework also maintained strong privacy (membership inference < 0.52) and low latency (~45 ms). SHAP analysis identified heart rate variability and electrodermal activity as key stress indicators. Overall, this work demonstrates a balanced approach to accuracy, privacy, efficiency, and interpretability in wearable health monitoring systems.

The rapid evolution of Industry 4.0 and emerging Industry 5.0 paradigms has accelerated the adoption of autonomous cyber-physical systems (CPS) in smart manufacturing, creating a need for robust runtime policy orchestration to ensure governance, regulatory compliance, and resilience under dynamic disruptions and cyber threats. This study addresses the lack of unified frameworks by proposing the Policy Resilient Orchestrator (PRO), a simulation-driven architecture integrating machine-readable policy cards, multi-agent reinforcement learning using Proximal Policy Optimization (PPO), multi-objective optimization via NSGA-II, runtime verification, and digital twin-based validation. A quantitative, desk-based methodology utilizing synthetic and benchmark datasets was employed to model CPS environments without physical hardware. NSGA-II generated Pareto-optimal policy configurations, while PPO enabled adaptive runtime control supported by simplex switching for resilience. Monte Carlo simulations across normal, attack, and failure scenarios showed significant improvements, with Governance Efficiency (0.961), Resilience Index (0.993), and Downtime Reduction (0.988), all statistically outperforming baselines (p < 0.0001). PRO demonstrates a scalable, reliable solution for trustworthy smart manufacturing systems.

This study examines the emerging threat of agentic ransomware, a novel class of cyberattack that targets autonomous artificial intelligence (AI) systems embedded within operational technology (OT) and industrial control system (ICS) environments. As industrial infrastructures increasingly integrate AI-driven automation, existing cybersecurity models remain insufficient to address threats capable of manipulating both digital processes and physical operations. To address this gap, this study adopts a quantitative analytical approach using three open-access datasets: Cybersecurity and Infrastructure Security Agency (CISA) ICS vulnerability advisories, the MITRE ATT&CK for ICS framework, and the U.S. Department of Energy OE-417 disturbance reports. The analysis integrates vulnerability exposure assessment, attack technique network modeling, and system reliability evaluation using frequency analysis, co-occurrence centrality metrics, and reliability indices. Findings reveal that industrial control devices account for 43.8% of identified vulnerabilities, while AI-driven analytics systems exhibit the highest severity (CVSS = 8.32) and significant remote exploitability (0.73). Furthermore, cyber-related disturbances demonstrate substantially lower reliability (0.28) compared to operational failures (0.83), indicating increased system instability under cyberattack conditions. Building on these findings, the study proposes and operationalizes an Agentic Ransomware Resilience Framework, introducing a composite resilience index that integrates vulnerability exposure, attack pathway centrality, and system recovery performance. The results highlight the critical role of AI-enabled components as emerging attack surfaces and emphasize the need for resilience-oriented cybersecurity strategies. The study concludes by recommending adaptive security architectures, network segmentation, AI-aware anomaly detection systems, and enhanced recovery mechanisms to safeguard intelligent industrial infrastructures against next-generation ransomware threats.

This research addresses governance and security gaps in autonomous artificial intelligence systems deployed within distributed cloud healthcare environments by proposing an integrated framework combining AI-driven governance, adaptive certification, behavioral authentication, and AI-agent risk monitoring within zero-trust architectures. Using a design science research methodology, three frameworks were developed and validated through simulation experiments, public healthcare datasets, and synthetic conversational AI traces, optimized for home-based resource-constrained setups. The adaptive certification lifecycle framework achieved a 4.7/5.0 compliance score against the NIST AI Risk Management Framework, with 94% latency reduction. The behavioral authentication framework, leveraging Isolation Forest and Mahalanobis distance, attained 94.2% F1-score and 1.2% false positive rate for non-human identity verification. The AI-agent risk monitoring framework, utilizing autoencoders and Q-learning, achieved 93.7% anomaly detection accuracy with 97.6% AUROC. Integrated multi-domain healthcare validation demonstrated 96.1% overall performance, reducing mean time to detection from 72.3 to 0.8 hours. These results highlight that continuous verification and adaptive mechanisms within zero-trust infrastructures substantially enhance security, regulatory compliance, and operational efficiency, providing a novel, quantifiable approach for resilient AI governance in healthcare.