In most traditional manufacturing industries, ensuring the quality of products or workpieces is crucial, beyond the processes of research and production. Through methods such as visual inspection, instrument measurements, pressure testing, as well as initial checks and subsequent inspections, regardless of how many checkpoints the product goes through within the factory, the final product delivered to the customer must be in a fully functional, defect-free, and high-quality condition! However, human attention is limited, and in the repetitive process of operating instruments for measurements and interpreting values, there may inevitably be oversights. When dealing with a large quantity of products, in order to optimize human resources, only sampling is often implemented. The sensitivity to defects can only be assured through statistical methods, without effective means of further improvement.

To assist the quality control needs in the manufacturing industry, the quality prediction system developed by MiTAC employs sensors installed at key points on processing machines to receive vibration data during the production process. This system records the behavior of machine components that cannot be observed externally. The collected data is then instantly analyzed and predicted using AI algorithms, reducing the burden of human observation and measurements. Finally, the system presents the quality level or health status of the product through a concise visual interface, achieving the goal of supporting quality management and preventing batch losses caused by the discovery of defects at a later stage.

MiTAC Quality Prediction System has been successfully deployed at TBI MOTION TECHNOLOGY CO.,LTD. as of June 2023. By collecting data from the machining processes of linear slides and sliders and integrating the necessary parallel accuracy information for quality analysis, a CNN model has been trained. This achievement enables real-time prediction of the grade of linear slides and sliders during the production process, ensuring that defects are promptly identified and establishing the first line of defense in quality control. Additionally, the system integrates measurement instruments such as depth gauges, automating the recording and management of data to contribute to the achievement of digital transformation goals.