Autonomous Artificial Intelligence Agents for Fault Detection and Self-Healing in Smart Manufacturing Systems
Akinde Michael Ogunmolu, Oluwaseun Oladeji Olaniyi, Anuoluwapo Deborah Popoola, Anthony Obulor Olisa, Oluwatobi Bamigbade
Journal of Energy Research and Reviews · pp. 20–37 · Published 9 Jul 2025
10.9734/jenrr/2025/v17i8445Abstract
This research develops and validates autonomous artificial intelligence agents for fault detection and self-healing in smart manufacturing systems within the context of Industry 4.0, addressing critical challenges in operational efficiency and reliability. Motivated by the need to reduce unplanned downtime and enhance production resilience, the study pursues three objectives: designing a robust AI architecture for fault detection, implementing effective self-healing mechanisms, and evaluating performance through rigorous testing. The literature review highlights gaps in temporal-spatial data integration and explainable self-healing, underscoring the potential of spiking neural networks (SNNs) and symbolic rule engines. Adopting a design science research paradigm, the proposed hybrid AI framework integrates SNNs, symbolic reasoning, and Isolation Forest algorithms, supported by industrial IoT sensor networks capturing vibration, thermal, and acoustic data. Validation was conducted through hardware-in-loop simulations, fault injection testing, and field deployments in automotive and electronics manufacturing environments. Results demonstrate a 97.3% fault detection accuracy and 89.4% self-healing recovery rate, reducing mean-time-to-repair by 31.7% and improving overall equipment effectiveness by 6.7 points, offering high annual savings for mid-sized manufacturers. Limitations include a 2.7% error margin in detection accuracy and an 83-second material regeneration latency, which exceed some high-speed production takt times. The research concludes by recommending adoption of the framework for improved manufacturing resilience and suggests future work in transfer learning and energy-efficient algorithms to enhance cross-domain scalability and sustainability.
Cited by 17
Bei-Qin Hu · Discover Computing · 2026
Shtwai Alsubai, Ahmad S. Almadhor, Abdullah Al Hejaili · Concurrency and Computation · 2026
K. Devi, Shangara Narayanee N. P, A. Anuradha · 2026 International Conference on Computational Robotics, Testing and Engineering Evaluation (ICCRTEE) · 2026
Zuhong Liu, Xiaohan He, Yubin Xie · International Journal of Industrial Ergonomics · 2026
G. Bharathi, Salini Suresh · 2026 Fourth International Conference on Augmented Intelligence and Sustainable Systems (ICAISS) · 2026
P. Latha, Suneel Pappala, S. Sree · 2026 5th International Conference on Communication, Computing and Electronics Systems (ICCCES) · 2026
Ali Zolfagharian, Liuchao Jin, Qi Ge · Advanced Intelligent Systems · 2026
Wei Gong, Shuang Qiao, Chenhong Cao · Journal of Industrial Information Integration · 2026
Guojun Wang · Discover Applied Sciences · 2025
A. Annan · International Journal of Scientific Research in Computer Science Engineering and Information Technology · 2025
Showing 13 of 17 known citations — external sources report more than can currently be individually listed.
Related research
- Leveraging Deep Learning Algorithms for Predicting Power Outages and Detecting Faults: A Review — shares topic coverage
- Error Analysis and Detection Procedures for Signature and Authentication Schemes — shares topic coverage
- Hybrid Approach for Fault Detection and Classification in Power Distribution Systems Using DWT and PSO Based-SVM with Real-Time Simulation — shares topic coverage
- Enhancing Power System Reliability with Machine Learning-enabled Fault Management — shares topic coverage
- Machine Learning Algorithm for Modelling and Analysis of Faults in Secondary Power Distribution Networks — shares topic coverage
Article metrics
Real usage data collected on this platform.
0
Page views
0
PDF downloads
0
Outbound clicks
17
Citations
Views by country
Approximate, from request IP at view time — not citizenship or institution. Countries with fewer than 5 views are grouped as "Other".
No views recorded yet.
Traffic sources
Referring site, by host.
No traffic recorded yet.
Views and downloads exclude known bots/crawlers. Citations combines this platform's own DOI-resolved index with each external source's own reported total — see Cited by above for individually listed citing works. Last refreshed 0 seconds ago.