Brief: Build a continuous CAD/mesh morphing pipeline to complement an ML surrogate used for external vehicle aerodynamics, enabling continuous design-space exploration beyond discrete part swaps.

Goals and tasks:

• Implement geometry morphing (FFD, RBF, ARAP or mesh-based parametric edits using Open3D/trimesh/PyVista) and ensure watertight, mesh-clean outputs within the surrogate’s training distribution.
• Define morphing parameters (body shape, ride height, frontal area, local control points, etc.).
• Generate batches of morphed geometries, run them through the surrogate, and analyze predicted aero metrics (Cd, Cl…).
• Produce a “morph gallery” showing sensitivity of aero metrics to geometric changes.
• Document a reusable morphing API for integration by the PFE student.

Profile/skills:

• Python + ML fundamentals (PyTorch/TensorFlow), 3D geometry processing experience, familiarity with Open3D/trimesh/PyVista/meshio.
• Basic CFD/aerodynamics understanding is helpful; exposure to mesh morphing or 3D learning is a plus.

Duration & perks:

• Target: students entering final year and interested in a later PFE.
• Period: 3 months.
• Paid: Monthly stipend.