Diagnosing AGV Wi-Fi Disconnection via dmesg Logs: A Real-World Factory Case Study

초록

Automated Guided Vehicles (AGVs) are fundamental to smart factory operations, yet maintaining stable Wi-Fi connectivity remains a significant challenge, as disconnections disrupt workflows and reduce productivity. We present a log-centric diagnosis method that parses kernel (dmesg) logs from production AGVs, extracts Wi-Fi-related events, and classifies event sequences into patterns. Analyzing a 3.5-month dataset from an automotive manufacturing factory, we identify eight recurring disconnection patterns and map each to its proximate cause. The resulting pattern-to-cause mapping provides practical insights for diagnose and mitigation. To our knowledge, this is the first systematic study to diagnose AGV Wi-Fi disconnections using kernel logs. Our method and findings offer a practical guideline that other plants can apply to address Wi-Fi disconnections, enabling faster diagnosis and more effective remediation of AGV connectivity issues in industrial environments.

키워드