In-situ monitoring and correction of geometric deviations in material extrusion additive manufacturing

Purpose This study aims to address layer shifting in Material Extrusion (MEX) additive manufacturing, a representative defect that compromises part quality and efficiency. Unlike previous studies that focus only on error detection, this research develops a complete, real-time in-situ monitoring and correction system capable of detecting and correcting geometric deviations under different error scenarios. The system improves precision, reliability and sustainability in MEX processes by reducing failed prints and material waste while maintaining geometric integrity. Design/methodology/approach A single-camera system captured contour images at checkpoint layers, and image processing algorithms detect geometric deviations. A novel correction strategy balanced overcorrection and undercorrection, ensuring robust error recovery. The system’s effectiveness was validated through systematic testing across multiple error magnitudes and positions, confirming its adaptability and performance under various shifting conditions. While the initial validation was performed on a square geometry for its simplicity and efficiency in demonstrating the concept, the approach was designed to serve as a foundation for broader application to more complex cross-sectional contours in future studies. Findings The system provides a complete real-time solution for detecting and correcting layer shifting errors, a capability not previously achieved. The detection algorithm exhibited absolute errors from 0.01 mm to 0.97 mm, demonstrating great prediction accuracy. The correction strategy effectively preserved part geometry despite deliberately introduced errors. Although monitoring increased printing time by 36%, the tradeoff is justified given the reduction in machine downtime and material waste. Originality/value This study introduces a cost-effective, real-time monitoring and correction system for MEX, bridging a critical gap in quality assurance by providing both detection and correction across varied error conditions. The findings lay the foundation for self-correcting, high-precision MEX processes, with potential applications in aerospace and healthcare, where geometric accuracy is crucial. By enabling scalable in-situ monitoring and correction, this system supports broader industrial adoption of additive manufacturing.