TOWARDS A DIGITAL TWIN FOR REAL TIME VISION BASED DIMENSIONAL INSPECTION AND STATISTICAL ANALYSIS

The research focuses on the integration of vision-based inspection and statistical process control enhanced dimensional quality control in additive manufacturing. The study utilized OpenCV and image-sensing tools to construct a non-contact inspection system. Dimensional features of 3D-printed parts underwent real-time measurement under optimized imaging conditions. The design of experiments framework controlled the lighting and part orientation, enabling consistent contour detection. Dimensional parameters extracted through OpenCV undergo evaluation using control limits to determine variation in process performance. Upper and Lower Control Limits define acceptable thresholds to prevent out-of-specification output during production. Analysis of collected dimensional data applied X-bar and R control charts to monitor trends and detect deviations linked to machine parameter shifts. Visualization of the Analysis data is through a Unity-based dashboard supporting remote display of inspection data. The system architecture contributed to the foundational Digital Twin development for predictive quality control and closed-loop process optimization.