The SoftGlove: A Tool for Posture Reconstruction of Robotic Hands in Challenging Scenarios

This paper presents a waterproof sensorized glove, the SoftGlove, intended to reconstruct the posture of a robotic hand even in conditions of low-visibility as severe as in the underwater environment. The reconstruction strategy implemented by the SoftGlove is based on a set of inertial motion units (IMUs) and on a complementary filter able to compute the relative orientation between two or more IMUs. The quality of the reconstruction is visualized for a full soft robotic hand wearing the SoftGlove. It is shown how this measurement system gives significant results in terms of accuracy, and its application affordability in many structures due to its mechanical robustness, small sizes of the employed sensors and reduced computational cost. Finally, examples of use of the system are reported in underwater environments.