We present Strain2Force, a compact nail-mounted sensing system that continuously estimates fingertip contact forces through nail surface strain. The device integrates four strain gauges positioned at the four corners of the nail to capture localized strain variations. We examine placement strategies under normal and shear loading, corresponding to an average surface strain of approximately 0.02% per Newton of applied normal force. A machine learning regression model maps multichannel strain responses to three-dimensional fingertip forces, with high accuracy across all components (Fx, Fy, and Fz; R2 = 0.877, 0.879, and 0.955; MSE = 0.028, 0.048, and 0.146). These results highlight a compact nail-based system capable of continuous, high-fidelity 3DOF force estimation through surface strain sensing.

Authors: Yanjun Chen, Xiping Sun, Jung-Hwan Youn, Craig Shultz

Best Paper Award Finalist!