Metallic object detection using passive magnetometer arrays for conveyor belt-based sorting systems

Object detection plays a crucial role in automation systems, such as in tasks where sorting of workpieces from conveyor belts is needed. Many technologies can be utilized in such systems for the given task. In this work, a method based on an array of passive magnetometers is proposed for the detection of metallic objects, which utilizes the generated magnetic disturbances. The advantages of this technology compared to existing ones can be the low price and the additional information that can be extracted from the signals. A measurement system was constructed for data acquisition. Measurements were collected using multiple objects for signal analysis to explore the possibilities of the system. Using the collected measurement data, a standard deviation-based algorithm was developed for the detection of workpieces. The proposed algorithm was tested in a determined task where two objects had to be detected and classified in multiple lanes of the conveyor belt. The results show above 86% detection efficiency, while the errors are mainly caused by the misclassification of the object type. The results obtained show the applicability of the proposed method for metallic object detection on conveyor belts.