TETSUO

The project targets the design and manufacturing of transtibial sockets which are lighter, stronger, and more durable by combining natural hemp fibers with recycled PET (rPET) via infrared-assisted filament winding. The aim is to automate the full “scan-to-socket” chain, from limb scanning to dimensional verification, reducing cycle time and variability. A multiphysics digital twin (ray tracing + heat conduction + crystallization + mechanics) will be developed to optimize process parameters and predict performance/defects. Machine-learning surrogates (e.g. BNNs) will speed up predictions and enable closed-loop control. The program spans materials characterization, modeling, experimental validation, and a TRL 3–5 demonstrator.

The Centre for Materials and Processes (CERI MP) of IMT Nord Europe leads the workpackages dedicated to the experimental validation and process monitoring, and to the manufacturing of a demonstrator, deploying an instrumented infrared-assisted filament winding bench (with thermal cameras, thermocouples) to map thermal fields, coalescence, and crystallinity. The CERI MP team is in charge of the development of the model–measurement correlation workflow (parameter calibration/uncertainty) and contributes to the microstructural and dimensional assessments (tomography/3D scanning) and the advanced mechanical testing. It integrates outputs from the digital twin and machine learning models to preemptively establish process windows online. IMT Nord Europe also drives dissemination (MOOC, exhibitions, papers) and supervises a PhD student on artificial intelligence-driven process control.