A Sensor and Machine Learning-Based Sensory Management Recommendation System for Children with Autism Spectrum Disorders

Abstract

Sensory processing issues are one of the most common issues observed in autism spectrum disorders (ASD). Technologies that could address the issue serve a more and more important role in interventions for ASD individuals nowadays. In this study, a sensory management recommendation system was developed and tested to help ASD children deal with atypical sensory responses in class. The system employed sensor fusion and machine learning techniques to identify distractions, anxious situations, and the potential causes of these in the surroundings. Another novelty of the system included a sensory management strategy making a module based on fuzzy logic, which generated alerts to inform teachers and caregivers about children’s states and risky environmental factors. Sensory management strategies were recommended to help improve children’s attention or calm children down. The evaluation results suggested that the use of the system had a positive impact on children’s performance and its design was user-friendly. The sensory management recommendation system could work as an intelligent companion for ASD children that helps with their in-class performance by recommending management strategies in relation to the real-time information about the children’s environment.