A Control Action Selection Model With an Explanation Mechanism for an Intelligent Automated 3d Printing Control System
Year: 2026
Type: article
Abstract: The paper addresses the problem of improving the validity, transparency, and practical usefulness of control interventions in intelligent automated FDM/FFF 3D printing systems. Modern approaches to monitoring additive manufacturing processes based on machine learning, computer vision, and multimodal telemetry analysis enable timely detection of defects, anomalies, and process deviations. However, in most cases they do not provide a sufficiently clear explanation of why a particular control action has been selected. As a result, operator trust in automated systems decreases, the analysis of false triggers becomes more difficult, and the effective use of intelligent control in laboratory, educational, and applied production environments remains limited. A control action selection model with an explanation mechanism is proposed. The model integrates the results of visual analysis of the print zone, printer telemetry indicators, G-code execution context, process phase, and integral risk level into a unified decision-making structure. Unlike traditional approaches, the proposed model not only generates decisions related to print continuation, pause, speed correction, temperature adjustment, or material feed adjustment, but also produces short and detailed explanations of the reasons for such interventions. The paper presents a formal description of the model, defines the set of possible control actions, describes the rules for mapping defect features to control actions, introduces an integral interpretability indicator, and provides a correspondence table between typical anomalies, interventions, and explanation texts. The practical value of the work lies in the possibility of using the model as an extension of existing intelligent supervisory 3D printing control systems. The proposed approach makes it possible to reduce false interventions, improve consistency between machine learning results and control actions, and increase operator trust in automated process control.
Language: English
Subfield: Industrial and Manufacturing Engineering
Field: Engineering
Domain: Physical Sciences
Open Access status: diamond
OpenAlex ID: W7147615465