Hybrid Edge-IoT Framework for Fault Detection and Reliability Enhancement in Smart Grids

Abstract

The increasing deployment of Internet of Things (IoT) devices and intelligent sensing technologies within modern smart grids has significantly improved grid visibility, operational automation, and real-time energy management capabilities. However, the growing complexity of smart grid infrastructures introduces critical challenges related to fault detection, communication latency, system reliability, and rapid decision-making. Traditional cloud-centric architectures often experience delays in processing large volumes of grid data, limiting their ability to respond effectively to time-sensitive operational events and fault conditions. The proposed framework integrates distributed IoT sensing devices, edge computing nodes, intelligent fault detection mechanisms, and centralized cloud services to enable real-time monitoring and localized decision-making. Operational data generated by smart meters, sensors, substations, and distributed energy resources are processed at nearby edge nodes, reducing communication overhead and enabling rapid fault identification. Furthermore, an intelligent reliability assessment module continuously evaluates grid conditions, communication performance, and equipment health to support proactive maintenance and resilient grid operation. The proposed framework enhances fault detection accuracy, minimizes response time, reduces network congestion, and improves overall grid reliability. Experimental evaluation under diverse smart grid operating scenarios demonstrates significant improvements in fault detection performance, communication efficiency, operational resilience, and system scalability compared with conventional monitoring architectures.

Introduction

The text presents a Hybrid Edge-IoT Framework for Fault Detection and Reliability Enhancement in Smart Grids, designed to address the challenges of real-time monitoring, fault detection, communication latency, and reliability in modern smart grid environments.

Smart grids increasingly rely on IoT devices such as smart meters, sensors, phasor measurement units (PMUs), intelligent electronic devices (IEDs), and distributed energy resources to generate large volumes of operational data. While these technologies improve grid monitoring and automation, traditional cloud-based architectures often suffer from communication bottlenecks, network congestion, high latency, and delayed fault responses, which can compromise grid stability and reliability.

To overcome these limitations, the proposed framework integrates IoT sensing, edge computing, intelligent fault detection, reliability assessment, and cloud services into a unified architecture. Instead of sending all data to centralized cloud servers, operational data is processed locally at edge nodes deployed near substations, renewable energy resources, and critical infrastructure. This enables rapid fault identification, localized decision-making, reduced communication overhead, and improved system responsiveness.

Key Contributions

1. Hybrid Edge-IoT architecture for real-time monitoring and localized processing of smart grid data.
2. Intelligent fault detection mechanism capable of identifying abnormal grid conditions and equipment failures with minimal delay.
3. Reliability enhancement model that continuously evaluates grid health, communication performance, and operational stability.
4. Comprehensive performance evaluation demonstrating improvements in fault detection accuracy, response time, communication efficiency, scalability, and overall reliability.

Literature Review Findings

Previous studies explored:

• Edge computing for fault detection,
• Blockchain and federated learning for decentralized monitoring,
• Digital twins for grid management,
• AI-based fault diagnosis,
• Predictive maintenance,
• Cybersecurity and resilience mechanisms.

However, most focused on specific aspects such as fault detection, security, or communication efficiency without integrating fault management, reliability assessment, adaptive control, and scalability into a single framework.

Proposed Methodology

The framework consists of five major stages:

1. IoT-Based Monitoring and Data Acquisition

• Collects real-time data from smart meters, PMUs, sensors, transformers, substations, renewable energy systems, and storage devices.
• Monitors voltage, current, frequency, power quality, equipment health, and communication status.
• Performs local data filtering and prioritization to reduce network traffic.

2. Edge-Intelligent Processing and Fault Detection

• Edge nodes process operational data near the source.
• Detect faults using parameters such as voltage deviations, frequency instability, transformer loading, communication failures, and temperature anomalies.
• Reduces fault detection latency through localized analysis.

3. Intelligent Fault Classification and Reliability Assessment

• Classifies faults according to severity and impact.
• Evaluates reliability using indicators such as:
  • Equipment health,
  • Communication reliability,
  • Network availability,
  • Power quality,
  • Service continuity.
• Generates reliability scores for proactive maintenance planning.

4. Adaptive Grid Control and Cloud Integration

• Initiates corrective actions including:
  • Load redistribution,
  • Fault isolation,
  • Demand response,
  • Distributed generation coordination,
  • Equipment protection.
• Cloud services handle long-term analytics, predictive maintenance, and system optimization.

5. Performance Evaluation

The framework is evaluated using:

• Fault detection accuracy,
• Response time,
• Communication latency,
• Throughput,
• Reliability index,
• Communication overhead,
• Scalability,
• Resource utilization efficiency.

Experimental Setup

The simulation included:

• More than 1500 interconnected smart grid devices,
• Smart meters, PMUs, IEDs, transformers, substations, renewable energy resources, battery storage systems, and sensors,
• Various fault scenarios such as equipment failures, voltage abnormalities, communication disruptions, and sensor malfunctions.

Results and Performance Analysis

Comparative Fault Detection Performance

The proposed framework achieved the highest fault detection accuracy and reliability while generating the lowest communication overhead.

Latency and Response Performance

The edge-based architecture significantly reduced latency and response time while increasing throughput.

Scalability Performance

Even as the number of connected devices increased from 500 to 6000, the proposed framework maintained reliability above 96.9%, outperforming cloud-based and conventional edge-assisted solutions.

Conclusion

This paper presented a Hybrid Edge-IoT Framework for Fault Detection and Reliability Enhancement in Smart Grids that integrates IoT-enabled monitoring, edge-intelligent data processing, fault identification mechanisms, reliability assessment, and adaptive grid management to improve the performance and resilience of modern smart grid infrastructures. The proposed framework enables localized processing of operational data at distributed edge nodes, thereby reducing communication delays and supporting rapid fault detection and response. Experimental results demonstrated the effectiveness of the proposed approach, achieving a fault detection accuracy of 98.1%, reliability index of 97.4%, and communication overhead of only 9.8%, outperforming conventional cloud-based and existing edge-assisted monitoring architectures. Furthermore, the framework attained the lowest average latency of 49 ms, highest throughput of 97.9 Mbps, and fastest response time of 58 ms, highlighting its suitability for real-time smart grid applications. The scalability analysis further revealed that the proposed framework maintained a reliability level of 96.9% even with 6000 connected devices, confirming its capability to support large-scale smart grid deployments. These results demonstrate that the integration of edge computing and IoT technologies can significantly enhance fault detection performance, communication efficiency, operational reliability, and overall grid resilience. Future research can focus on incorporating Artificial Intelligence and Deep Learning models for predictive fault forecasting, integrating blockchain-based security mechanisms for secure edge communication, developing federated learning-enabled distributed intelligence across edge nodes, supporting autonomous self-healing grid operations, and extending the framework to manage renewable energy-rich smart grids, electric vehicle charging infrastructures, and next-generation cyber-physical energy systems.

References

[1] T. Singh et al., \"Intelligent FaultEdge: AI-Driven Fault-Tolerant Edge Framework for Smart Grid Monitoring in IoT,\" 2024 IEEE 12th International Conference on Intelligent Systems (IS), Varna, Bulgaria, 2024, pp. 1-6, doi: 10.1109/IS61756.2024.10705247. [2] K. Senthilkumar and P. Dass, \"Fault Detection and Isolation in Smart Grids Using IoT and Edge Computing,\" 2025 2nd International Conference on Artificial Intelligence and Knowledge Discovery in Concurrent Engineering (ICECONF), Chennai, India, 2025, pp. 1-8, doi: 10.1109/ICECONF65644.2025.11379432. [3] K. J. Mukisa, L. A. C. Ahakonye, J. M. Lee and D. -S. Kim, \"Synergistic Smart Grid: Instability and Fault Detection Using Blockchain and Federated Learning,\" 2025 Sixteenth International Conference on Ubiquitous and Future Networks (ICUFN), Lisbon, Portugal, 2025, pp. 588-593, doi: 10.1109/ICUFN65838.2025.11170004. [4] D. Sarathkumar et al., \"Digital Twin Technology for Smart Grid Monitoring and Control,\" 2026 IEEE International Students\' Conference on Electrical, Electronics and Computer Science (SCEECS), Bhopal, India, 2026, pp. 1-6, doi: 10.1109/SCEECS68810.2026.11429733. [5] Parasdeep, Ayushi, D. Upadhaya, M. Pal, M. Kumar and S. Singh, \"AI-Driven Smart Grid Fault Diagnosis Using Federated Learning and Edge Computing in IoT-based Power Systems,\" 2025 IEEE 7th International Conference on Computing, Communication and Automation (ICCCA), Greater Noida, India, 2025, pp. 1-6, doi: 10.1109/ICCCA66364.2025.11325716. [6] R. T. Balla, S. Padmanabham, R. S. Bhandari, V. Kumar and S. Khattar, \"Scalable IoT Frameworks for Smart Grid Management: A Time-Series and Predictive Maintenance Analysis Using ARIMA Models,\" 2026 5th International Conference on Innovative Practices in Technology and Management (ICIPTM), Noida, India, 2026, pp. 1-6, doi: 10.1109/ICIPTM69057.2026.11465901. [7] R. Kant, S. Sharma, V. Vikas, S. Chaudhary, A. K. Jain and K. K. Sharma, \"Blockchain – A Deployment Mechanism for IoT Based Security,\" 2023 International Conference on Computational Intelligence, Communication Technology and Networking (CICTN), Ghaziabad, India, 2023, pp. 739-745, doi: 10.1109/CICTN57981.2023.10140715. [8] M. Y. H. Mjbas, S. A. Aliesawi and M. R. H. Al-Dahhan, \"Deployment-Aware Benchmarking of Lightweight Machine Learning for IoT-Based Smart Grid Fault Detection: A Comparative Study Across Easy, Noisy, and Public Evaluation Settings,\" 2026 8th International Congress on Human-Computer Interaction, Optimization and Robotic Applications (ICHORA), Ankara, Turkiye, 2026, pp. 1-9, doi: 10.1109/ICHORA69329.2026.11537190. [9] M. Y. Masood, H. Rifà-Pous, M. Awais, A. Khawar and M. Muzaffar, \"Co-Simulation of FDI and DOS Attacks in IoT-Enabled Next-Generation Smart Grids: A Framework for Resilience,\" 2026 7th International Conference on Advancements in Computational Sciences (ICACS), Lahore, Pakistan, 2026, pp. 1-7, doi: 10.1109/ICACS69208.2026.11433141. [10] V. Sharma and S. Kumar, \"Role of Artificial Intelligence (AI) to Enhance the Security and Privacy of Data in Smart Cities,\" 2023 3rd International Conference on Advance Computing and Innovative Technologies in Engineering (ICACITE), Greater Noida, India, 2023, pp. 596-599, doi: 10.1109/ICACITE57410.2023.10182455. [11] L. Zeng, Y. Shi, L. Li, J. Zhu and R. Tan, \"Research on the Application of Knowledge Management System and Digital Service in Smart Grid,\" 2024 2nd International Conference on Mechatronics, IoT and Industrial Informatics (ICMIII), Melbourne, Australia, 2024, pp. 527-531, doi: 10.1109/ICMIII62623.2024.00104. [12] A. S. Dhinakaran, E. Harish, P. Maniraj, R. V. Kamali, R. Vishwangar and G. Devayani, \"AIoT-Enabled Smart Flood Detection and Autonomous Power Isolation System for Real-Time Safety and Predictive Risk Management,\" 2026 4th International Conference on Intelligent Data Communication Technologies and Internet of Things (IDCIoT), Ottapalam, India, 2026, pp. 286-291, doi: 10.1109/IDCIoT67589.2026.11455663.

Copyright

Copyright © 2026 Saba ., Sharad Kumar. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.