Assistance app for a humanoid robot and digitalization of training tasks for post-stroke patients

Abstract

During the last decades the number of people affected by stroke increased a lot. Unfortunately, the number of therapists is not large enough to fulfill the demands for specific training for stroke survivors. Within the project E-BRAiN (Evidence-based Robot Assistance in Neurorehabilitation), we want to develop software that allows a social humanoid robot to give instructions to perform and to observe carefully selected exercises and provide feedback. In addition, the robot should motivate patients in continuing their training tasks. We focus in this paper on the Arm Ability Training (AAT). Conventionally, some AAT exercises are performed with paper and pencil under the supervision of therapists. To improve the collaboration between patient and a humanoid the training tasks have to be digitalized. Such a digitalization of one training task is discussed by using a tablet computer. This allows the social humanoid robot to obtain information about the training performance. The robot can provide instructions, show pictures and videos. In addition, the humanoid can comment on the training results, can aid and motivate. Comments on the long-term results of the training tasks related to the individual goals of stroke survivors can be provided as well.