Skip to main content
. 2020 May 4;11:2183. doi: 10.1038/s41467-020-15990-7

Fig. 6. Orthogonal recognition of human grips for dynamic robotic manipulation.

Fig. 6

a Schematic shows the process of identifying human grip signatures for manipulating robotic grips. Both human and robotic grips involve two steps, namely the minFist and tightening. The coupled electromechanical patterns retrieved by CoupOn can orthogonally recognize the speed and strength of human grips, therefore allowing the categorization of four permutations in the gesture matrix, according to speed of the two steps. Scale bar, 2 cm. b Plots show the retrieved excitation–contraction signatures corresponding to the four permutations in gesture matrix. Data obtained on subject 5. c Photograph shows the setup for the robotic grip of a soft yellow ball with the ring finger connected with a force gauge. d Plots indicate the grip force exerted on the soft ball during the four permutations of recapitulated grips.