Comprehensive adaptive non-planar and planar slicing model for LMD-wire robotic additive manufacturing: Literature review, model development, and experimental validation

Robotic additive manufacturing using wire-laser directed energy deposition (LMD-wire) enables the fabrication of large-scale metal components with high geometric flexibility. However, existing slicing and toolpath generation solutions are often limited to planar strategies, proprietary software, or lack integration with industrial robotic systems. This work presents a validated open-source framework for planar and non-planar robotic additive manufacturing based on the LMD-wire process. The framework integrates adaptive slicing, geometry-driven tool orientation, process-aware velocity modulation, and direct robot code generation within the Rhino 3D and Grasshopper environment. Experimental validation was conducted in an industrial cell equipped with a KUKA KR70 R2100 robot and a Meltio Engine Robots deposition system. For planar slicing, the maximum dimensional manufacturing error estimated using the six-sigma (6 σ ) methodology was approximately 1 mm. For non-planar slicing, the maximum Hausdorff distance and Root Mean Square Error (RMSE) were approximately 2 mm and 0.9 mm, respectively. The results demonstrate stable deposition, accurate geometric reproduction, and improved adaptability for complex non-planar geometries. The proposed framework provides a reproducible and extensible open-source solution for advanced robotic metal additive manufacturing.