In the semiconductor manufacturing industry, efficient operation of wafer transfer robots has a direct impact on productivity and product quality. Ball screw misalignment anomalies are a critical factor affecting precision transport of robots. Early diagnosis of these anomalies is essential to maintaining system efficiency. This study proposed a method to effectively diagnose ball screw misalignment anomalies using 1D-CNN and 2D-CNN models. This method mainly uses binary classification to distinguish between normal and abnormal states. Additionally, explainable artificial intelligence (XAI) technology was applied to interpret diagnostic decisions of the two deep learning models, allowing users to convince prediction results of the AI model. This study was based on data collected through acceleration sensors and torque sensors. It compared accuracies of 1D-CNN and 2D-CNN models. It presents a method to explain the model"s predictions through XAI. Experimental results showed that the proposed method could diagnose ball screw misalignment anomalies with high accuracy. This is expected to contribute to the establishment of reliable abnormality diagnosis and preventive maintenance strategies in industrial sites.

A linear motor is an actuator that has strong thrust and high controllability, and can perform linear motion without the use of a motion converter. In this study, we propose a new method to measure the position of the mover of a permanent magnet linear synchronous motor by measuring the magnetic flux density. To resolve the problem that existing methods have to spatially arrange multiple sensors, the proposed method uses a two-dimensional magnetic flux density measurement value at one point. In accordance with this, the estimation method was modified, the convergence condition of the estimation method was obtained, and the time required for the calculation was estimated. The validity of the proposed method was verified through comparative experiments with existing methods. As a result of the test, the proposed method had a small maximum absolute error compared to the existing methods, and was robust against sensor gain changes.