A Robot-Assisted Scenario Training for Students with ASD

Students with autism spectrum disorders (ASD) often feel insecure in new environments due to social challenges, unfamiliarity, and a lack of support or understanding. Despite considerable efforts dedicated to assisting students in adapting to new environments and understanding appropriate behaviours in public settings, there remains a lack of interactive and personalized learning systems. In this work, we developed a robot-assisted scenario training (RAST) system to facilitate inclusive learning and arouse students' learning interests. With the RAST system, we seek to identify effective interactions that can improve students' engagement. To this end, we invited 13 students with ASD to participate in an evaluation study. In the study, self-determination theory (SDT) measures students' learning engagement. Learning engagement and effectiveness are evaluated using variance analysis (ANOVA). Students also participated in interviews to report their user experience regarding the system. The results reveal that learning with the RAST system can significantly arouse students' intrinsic motivation and improve their behavioural, emotional, and cognitive engagement. Additionally, students with ASD increased their learning performance by 8.33%. Furthermore, students exhibited a high level of engagement in scenario training with certain types of interactions, including personalized functions, visual cues and sound quality. Overall, the RAST system demonstrates promising capabilities in enhancing students' learning engagement and proficiency with ASD.