Experiments in the detection of upper limb posture through kinestetic strain sensors

Conductive elastomers are strain-sensing technology which can be embedded unobtrusively into a garment's fabric. A prototype was realized to simultaneously measure the strains at multiple points of a shirt covering the thorax and upper limb. This paper describes preliminary experiments with machine learning techniques, employed to analyse the strain measures in order to reliably reconstruct upper-limb posture. The scope of the application is to detect execution, correctness and progress of physical exercises performed as part of neurological rehabilitation therapy.