Interpretable AI Prediction of Melt Pool Geometry in IoT-Enabled Additive Manufacturing in Industry 4.0 Utilizing CatBoost: Decision-Making via NSGA-II

Additive manufacturing (AM), such as Laser Powder Bed Fusion (LPBF), enables complex geometries and custom components, with melt pool geometry control essential for quality. This paper employs CatBoost regression to optimize melt pool characteristics-depth, width, and length-based on process parameters in IoT-enabled AM under Industry 4.0. Using extensive datasets, the models effectively predict melt pool behavior, with CatBoost excelling in length prediction (R-squared: 0.9690). Integrating Multi-Objective Optimization (NSGA-II) enhances real-time monitoring, leveraging IoT for better decision-making and efficiency. This research advances the understanding of melt pool dynamics, improving AM quality, reliability, and digital manufacturing outcomes.