RESUMEN
Loss of an upper limb exerts a negative influence on an individual's ability to perform their activities of daily living (ADLs), reducing quality of life and self-esteem. A prosthesis capable of performing basic ADLs functions has the capability of restoring independence and autonomy to amputees. However, current technologies present in robotic prostheses are based on rigid actuators with several drawbacks, such as high weight and low compliance. Recent advances in robotics have allowed for the development of flexible actuators and artificial muscles to overcome the limitations of rigid actuators. Dielectric elastomer actuators (DEAs) consist of a thin elastomer membrane arranged between two compliant electrodes capable of changing dimensions when stimulated with an electrical potential difference. In this work, we present the design and testing of a finger prosthesis driven by two DEAs arranged as agonist-antagonist pairs as artificial muscles. The soft actuators are designed as fiber-constrained dielectric elastomers (FCDE), enabling displacement in just one direction as natural muscles. The finger prosthesis was designed and modeled to show bend movement using just one pair of DEAs and was made of PLA in an FDM 3D printer to be lightweight. The experimental results show great agreement with the proposed model and indicate that the proposed finger prosthesis is promising in overcoming the limitations of the current rigid based actuators.
RESUMEN
Lower-limb gait training (GT) exoskeletons have been successfully used in rehabilitation programs to overcome the burden of locomotor impairment. However, providing suitable net interaction torques to assist patient movements is still a challenge. Previous transparent operation approaches have been tested in treadmill-based GT exoskeletons to improve user-robot interaction. However, it is not yet clear how a transparent lower-limb GT system affects user's gait kinematics during overground walking, which unlike treadmill-based systems, requires active participation of the subjects to maintain stability. In this study, we implemented a transparent operation strategy on the ExoRoboWalker, an overground GT exoskeleton, to investigate its effect on the user's gait. The approach employs a feedback zero-torque controller with feedforward compensation for the exoskeleton's dynamics and actuators' impedance. We analyzed the data of five healthy subjects walking overground with the exoskeleton in transparent mode (ExoTransp) and non-transparent mode (ExoOff) and walking without exoskeleton (NoExo). The transparent controller reduced the user-robot interaction torque and improved the user's gait kinematics relative to ExoOff. No significant difference in stride length is observed between ExoTransp and NoExo (p = 0.129). However, the subjects showed a significant difference in cadence between ExoTransp (50.9± 1.1 steps/min) and NoExo (93.7 ± 8.7 steps/min) (p = 0.015), but not between ExoTransp and ExoOff (p = 0.644). Results suggest that subjects wearing the exoskeleton adjust their gait as in an attention-demanding task changing the spatiotemporal gait characteristics likely to improve gait balance.