Optical Microfiber Neuron for Finger Motion Perception

To achieve dexterous motion controlling of robot, the sensors that function like human neurons for motion perception are essential. In this work, a silica microfiber probe-based optical neuron (MPON) for robot finger motion detection is proposed. The silica microfiber probe was fabricated by snapping a biconical silica optical microfiber that drawn from the standard optical fibre. Then it was embedded into thin polydimethylsiloxane (PDMS) to detect and recognize motions of robotic finger. Specifically, a PDMS-Teflon-Microfiber-Teflon-PDMS composite structure was prepared to protect the waveguide structure of silica microfiber probe and avoid the environmental pollution. With the help of this composite structure, the proposed MPON achieved the accurate measurement of bending angle with large range and fast response. The repeatability and stability of MPON were also investigated. Additionally, different finger motions were successfully distinguished through observing the output power variation of MPON. The proposed MPON could serve as the perceptron of robot hand, which could be applied in dexterous gesture control even human machine interaction.

Automated summary

A silica microfiber probe-based optical neuron (MPON) for robot finger motion detection was proposed in this study. The use of a composite structure protected the waveguide structure of the silica microfiber probe and successfully distinguished different finger motions.