Origami Single-end Capacitive Sensing for Continuous Shape Estimation of Morphing Structures

The work introduces a sensing method called FxC for morphing structures made with folding (origami) patterns. Conductive materials are embedded into origami areas to form single-end capacitive sensing patches, and the authors report that the sensor signals change in a way that matches the structure’s motion. The authors compare behavior using 3D geometry simulation plus physics-based theoretical deduction, and they state that the simulated behavior is similar to what is observed experimentally. They then build a software pipeline that uses sensor signals to reconstruct the dynamic structural geometry using deep neural network regression of geometric primitives extracted from vision tracking. They test multiple folding patterns (Accordion, Chevron, Sunray, and V-Fold) with different sensor layouts using paper-based and textile-based materials. Reported results include up to 95% correlation (R-squared) between predicted geometry primitives and visual ground truth, with a tracking error of 6.5 mm for patches.