AoI-Driven IoT Fault Detection for Smart Grids
33rd IEEE Conference on Signal Processing and Communications Applications, SIU 2025, İstanbul, Turkey, 25 - 28 June 2025, (Full Text)
- Publication Type: Conference Paper / Full Text
- Doi Number: 10.1109/siu66497.2025.11112068
- City: İstanbul
- Country: Turkey
- Keywords: Age of Information (AoI), IoT Fault Detection, Real-Time Monitoring, Semantic Communication, Smart Grids
- Middle East Technical University Affiliated: Yes
Abstract
Ensuring the reliability and efficiency of electric distribution grids necessitates real-time fault detection systems that prioritize data freshness and transmission efficiency. This paper presents the first Age of Information (AoI)-driven IoT fault detection system, powered by an AoI-optimized algorithm that dynamically adjusts transmission intervals to ensure timely and critical updates. By integrating ESP32 and STM32 micro-controllers with an MQTT-based communication framework, the system adapts to network conditions, mitigating congestion while improving data precision. As a pilot project, the system was deployed across five transformer stations on a live power grid, where it successfully detected 27 fault types with high accuracy. Experimental results show timely fault detection with minimal data use - 17 KB/year in fault mode and 55 MB/year in continuous mode - enabling long-term, battery-powered operation.