Background: Monitoring body kinematics has fundamental relevance in several biological and technical disciplines. In particular the possibility to exactly know the posture may furnish a main aid in rehabilitation topics. In the present work an innovative and unobtrusive garment able to detect the posture and the movement of the upper limb has been introduced, with particular care to its application in post stroke rehabilitation field by describing the integration of the prototype in a healthcare service.

Textile-based transducers are an innovative category of devices that use conductive fibres meshed with elastic textile fabrics. Within this paper, a new class of strain sensors, which represents an excellent trade-off between figures of merit in mechano-electrical transduction and possibility of integration in textiles, is presented. Electrically conductive elastomer composites show piezo-resistive properties when a deformation is applied. Conductive elastomer can be applied to fabric or to other flexible substrate and they can be employed as strain sensors. We integrated conductive elastomer sensors into fabrics to realize wearable kinaesthetic garments able to detect posture and movement of a user. This paper deals with the design, the development and realization of a set of sensing garments, from the characterization of innovative textile-based sensors to the methodologies employed to gather information on the posture and movement from the entire garments. Data deriving from the prototypes are analysed and compared with those deriving from a traditional movement tracking system. The realized kinaesthetic garments have shown very promising performance in terms of body segment position reconstruction and posture classification.

This paper describes the design, the development and the preliminary testing of a wearable system able perform a real time estimation of the local curvature and the length of the spine lumbar arch. The system integrate and fuse information gathered from textile based piezoresistive sensor arrays and tri-axial accelerometers. E-textile strain sensing garments suffer from non-linearities, hysteresis and long transient, while accelerometers, used as inclinometers, present biased values and are affected by the system acceleration due to subject movements. In this work, focused on the wearability and comfort of the user, we propose a fusion of the information deriving from the two class of sensors to reduce their intrinsic errors affecting measurements. Comparative evaluation of system performances with stereophotogrammetric techniques shows a 2% error in lumbar arch length reconstruction.