Employing Explainable Artificial Intelligence (XAI) Methodologies to Analyze the Correlation between Input Variables and Tensile Strength in Additively Manufactured Samples

Transparency in manufacturing AI is achievable: This research demonstrates that explainable AI techniques like SHAP can be successfully applied to manufacturing processes, making it possible to understand why an AI system makes predictions about product quality rather than treating it as a "black box."

Not all input factors matter equally: The study found that Infill percentage and Extrusion Temperature are the primary drivers of tensile strength in 3D-printed objects, while Layer Height and Print Speed have relatively minor effects. This suggests that responsible AI implementation should focus monitoring and validation efforts on the most influential variables.

Complex relationships require appropriate tools: The researchers discovered that the relationship between manufacturing parameters and product strength is "highly intricate and nonlinear," meaning simple linear models cannot capture how these factors interact. This highlights why explainable AI methods are necessary—they can reveal hidden patterns that traditional analysis might miss.

XAI enables better decision-making in manufacturing: By using SHAP to interpret machine learning predictions, manufacturers can identify which specific input parameters most influence product quality, enabling more informed decisions about process optimization and quality control.

Responsible AI requires interpretability from the start: This paper demonstrates that explainability should be built into AI systems for manufacturing from the beginning, rather than added afterward, allowing stakeholders to understand and trust the factors driving production decisions.