Abstract:

Speech sound disorders refer to any difficulty with perception, motor production, or phonological representation of speech sounds and speech segments. Between 2.3% and 24.6% of school-aged children have a speech sound disorder. Technology provides a great opportunity to reduce and facilitate the assessment and intervention of these disorders. Particularly, robots are a valuable tool in education, psychology and rehabilitation.

The goal of this work is present a new project, denoted RO.MU.LO (RObot MUltifunctional for LOgopaedics), to develop a new technology-based solution to detect and treat children with speech sound disorders for speech and language therapists.

An ad-hoc cloud infrastructure will be developed to offer a series of capabilities. Such decision benefits the access to high-computation resources to train AI models. In this way, the proposed solution should provide different modules:
a) a module to evaluate patients;
b) a module to give a diagnosis of the test passed to the patient, and
c) a module to recommend a treatment for this patient.

RO.MU.LO aims to provide multiple benefits for speech and language therapy sessions providing an easy and fast assessment, a diagnosis and a personalized treatment for every child according to her or his specific needs.

This solution will reduce costs, increase quality, and make speech and language intervention more personalized for children with speech sound disorders. With a tool that works more autonomously and removes the delay between sessions to prepare the robot, the therapist will have it all in only one tool.

Keywords:

Intelligent system, speech disorders, robotics.