Abstract:
The present work is based on the implementation of a human-robot HRI interaction system to support the sessions of gross motor rehabilitation in infants. The proposal is formed by a humanoid robot, which indicates the exercise of recovery that the child must imitate through movements of their extremities. At the same time, the mechanism visually and auditorily motivates the child to ensure that the proposed exercises are carried out in a correct manner and thus assess the child's progress in therapy. A Field-Programmable Gate Array (FPGA) is used for the architecture of the system, which is responsible for controlling the robot as well as the validation algorithm and an auxiliary processor for the acquisition and processing of the information taken from the sensors in real time. For the validation of the exercises proposed in the therapy, the system proposes an algorithm based on classifiers. The algorithm is based on the comparison of patterns of the exercises performed by the humanoid in contrast to the acquired movements of inertial measurement sensors subject to the user.
