Thin-walled objects, such as sheet metals or casting parts, are modelled as mid-surface in Finite Element Analysis (FEA), where the extraction of mid-surface(mesh) is still problematic especially for objects with complicated contours. In addition, market demand for more variety of new designs required the modelling to be done in much shorter time, which makes it indispensable to improve the meshing efficiency. As an attempt to improve the modelling efficiency for thin-walled objects, we developed a robotic system to automatically retrieve mid-mesh from an input 3D-CAD data. This system has its own shape recognition engine and mesh-rule knowledge database that enables it to recognize shape, generate and control mesh configuration/pattern at the mid-surface position. Moreover, this will provide the users with controllability over mid-surface generation and meshing process for various features such as ribs, fillets, holes, clips, and many others. This tool has the following advantages: 1) Full automation with no human intervention, 2) Controllability over mid-surface generation, 3) Controllability over mesh patterns. Fully automated with unmanned operation allow it to run 24-hours non-stop thus will eventually improve the meshing efficiency. Through an embedded shape recognition engine, mid-surface generation process is controllable where various types of mid-surface shapes can be generated for one specific feature. Most of the time, the required shapes of mid-surface may vary between enterprises depends on their experience and expertise in CAE analysis. Depending on the analysis objectives, the required mesh patterns are also different between crash, strength, and thermal analysis. Users are allowed to select their analysis objective before the mesh generation process so that better results are possible. For future implementation, this tool will provide easy accessibility for design engineers to transform between CAD and CAE domain, and allow them to create reliable mesh models by their own, instead of relying on design analysts.