Teabot: Robô para treinamento de expressões faciais emocionais para pessoas com Transtorno do Espectro do Autismo

Abstract

Actually, there is one case of autism for every 110 people in the world, and it is estimatedthat in Brazil there are about 2 million individuals with autism. Autistic Spectrum Disorder ischaracterized by persistent deficits in social communication and social interaction in multiplecontexts; individuals with this disorder have difficulties in their emotional awareness, resultingin a condition that known as Alexithymia, characterized by the difficulty in identifying anddescribing feelings, this condition affects 50% of the autistic population. In this way, it ispossible to use treatments such as behavioral analysis, occupational therapies, speech therapist,physiotherapy, aimed to reducing the symptoms caused by this disorder. On the other hand, itis possible to see in the literature robots that is capable of interacting with autistic individuals,aiming at improving some specific skill of these people, these robots are called socially assistive.In this sense, this work aimed to develop an assistive social robot capable of training therecognition and representation of the six universal facial expressions in children and adolescentswith autism, and have difficulties in performing these expressions. The robot developed in thiswork have a non-humanoid characteristics, and still have audiovisual resources. All the physicaland logical components are controlled by a Raspberry Pi 3 and implemented a system to representthe six universal facial expressions, which incorporated the facial recognition system. Threedifferent versions of the robot were developed. Finally, two experiments were developed, witha total of 19 participants, in 19 days dedicated to the execution of the experiments, more than50 sessions were carried out with the participants, reaching a little more than 1000 minutes ofinteraction with the individuals and, captured more of 15 thousand images by two versions ofTEABot. Based on the analysis of the results obtained with the execution of the experiment,it became possible to identify that the participants presented greater difficulty in imitatingexpressions of sadness, anger, disgust and fear. According to the experiments carried out and theinformation collected with the parents of the participants, it became possible to gather evidencethat the TEABot robot is able to train facial expressions in people with difficulties in representingand recognizing these emotions.