A Machine Learning Approach for Navigation Quality Evaluation of Frequency Guided AGVs

The rise of artificial intelligence (AI) is significantly impacting industrial processes, including the automation of guided vehicle (AGV) navigation in logistics. A common AGV navigation method is frequency-based, using inductive loops embedded in the floor connected to a frequency generator. However, this system is prone to failures, such as loop degradation, generator or antenna malfunctions, and controller tuning issues, which can affect navigation quality. To address these challenges, this study aims to train and compare six machine learning models to identify the one with the best performance in classifying the navigation quality of frequency-guided AGVs. By identifying the most effective model, this research seeks to improve predictive maintenance and enhance the efficiency of AGV navigation systems, thereby increasing the reliability of industrial logistics operations.