The Impact of an In-Home Co-Located Robotic Coach in Helping People Make Fewer Exercise Mistakes

Abstract

The conventional smart gloves present a challenge regarding their portability as most work on gesture recognition techniques based on vision sensing and image processing. The multiple algorithms and signal filtering further make the overall process cumbersome. This work proposes a Shape Memory Alloy (SMA) integrated sensing mechanism in a smart glove for autonomous control. A novel hand gesture recognition technology is developed using kinaesthetic feedback from the finger joint movements. The paper presents a smart glove with an external SMA embedded tubing attachment for the thumb, index, and middle fingers. The motion of the SMA wires is constrained between a fixed end on the tip of the fingers, and the other end is connected to a linear position sensor with spring feedback. The SMA wires in this design exist in their Austenite phase at room temperature, thus exhibiting superelastic or pseudoelastic behavior. The tension in the SMA wire is observed and measured upon bending the fingers, corresponding to the mechanical travel in the linear position sensor. The individual and a combination of position sensor readings are then used as commands for actuating interactive toys. Using a three-finger approach, one can extract seven commands depending upon single or multiple finger movements. This data is further used to actuate the toys, and a use-case for cobotic application is proposed to help better understand interactive play, hand-eye coordination, and thus early cognitive development in children with Autism Spectrum Disorder (ASD). The discrete data output with binary data is independent of other devices or heavy data processing requirements, thus making the proposed novel SM-EXO a better alternative for non-portable and complex smart gloves.