Massachusetts Institute of Technology engineers have developed a flexible, shape-sensing silicone sheet embedded with custom rubber optical fibers that digitally reconstructs its 3D form in real time as it bends and twists, according to a study published in Advanced Intelligent Systems.
Traditional motion-capture systems rely on rigid sensors stitched into garments at fixed locations. To move past these stiff designs, researchers at the Massachusetts Institute of Technology turned to soft optics. Graduate student Qifan Yu and his colleagues designed a shape-sensing sheet that uses flexible waveguides to continuously track surface deformations.
Rubber Waveguides Enable Bidirectional Bending Detection
Standard optical fibers consist of a clear glass core surrounded by an opaque dark cladding, trapping light inside for efficient telecommunications data transmission. When bent, these standard lines allow less light to exit the far end, providing a basic measure of curvature.
For the new sensor, the research team fabricated their own waveguides out of transparent, flexible rubber. They dyed the surrounding cladding black and intentionally roughened just one side of the core. When the fiber bends toward its rough side, scattered light alters the total output differently than a bend toward the smooth side. This asymmetry turns each half-roughened strand into a bidirectional shape sensor.
Zig-Zag Fiber Layouts Improve 3D Shape Reconstruction
Before building physical prototypes, Nina Cao and fellow researchers simulated various embedded fiber layouts inside silicone sheets, ranging from straightforward checkerboards to criss-crossed, zig-zag patterns. By testing simulated folds and twists, the team converted light output measurements into bending estimates to reconstruct overall 3D shapes. A specific spacing of zig-zagging fibers proved closest to the original shapes.
The physical prototype integrates an LED at one end of each 1-millimeter-thick fiber and a light sensor at the other end. These components connect to an external circuit board that collects and amplifies the light measurements. An automated algorithm then translates those light readings into a virtual 3D reconstruction almost in real time.
In experiments, they showed that the algorithm smoothly created a virtual reconstruction of the sheet, almost in real time, as the researchers twisted it into different forms.
MIT News reporting
Soft Sheets Outperform Rigid Sensors in Accuracy
To test accuracy, the MIT team placed the soft sheet over various 3D-printed molds. Using a metric measuring the distance between the actual surface and the digitally reconstructed surface, they found their error was under 0.4 centimeters. By comparison, existing designs utilizing rigid sensors typically produce errors around 1 to 2 centimeters.
The design also proved resilient: even when a few fibers are cut or disconnected, the remaining strands still accurately reconstruct the overall shape of the sheet.
Looking ahead, the research team aims to shrink the optical fibers from 1 millimeter down to tens of micrometers—thinner than a human hair—using alternative fabrication processes. Assistant professor of mechanical engineering Kaitlyn Becker noted that embedding many of these thinned fibers into garments could help physical therapists track patient movement and rehabilitation progress over extended periods.
These sensing sheets could help a physical therapist to recall and compare more precisely how an evaluation went today versus two months and hundreds of appointments ago.
Kaitlyn Becker, assistant professor of mechanical engineering at MIT