This study extends the framework proposed by Campitiello et al. (2024) on the educational robot with AI, focusing on emotion recognition and social skills in children with autism spectrum disorder (ASD). The research explores additional key aspects according to Howlin et al.’s (1999) socio-cognitive development program. Specifically, it addresses the recognition of schematic designs as social icons and simplified facial expressions, as well as the understanding of desires and intentions through computational models of Theory of Mind. The system architecture utilizes neural networks, tools particularly effective in visual pattern recognition. This study opens new perspectives for supporting children with ASD in acquiring social skills in everyday contexts. However, the main challenge lies in ensuring that the proposed methods can dynamically adapt to individual needs, with particular attention to ethical considerations and the sustainability of the adopted solutions.
Il presente studio amplia il framework di Campitiello et al. (2024) sul robot educativo con IA incentrato sul riconoscimento delle emozioni e sulle abilità sociali in bambini con disturbo dello spettro autistico (ASD). La ricerca esplora ulteriori aspetti chiave secondo il programma di sviluppo socio-cognitivo di Howlin et al. (1999). In particolare, il riconoscimento di disegni schematici come icone sociali ed espressioni facciali semplificate, e la comprensione di desideri e intenzioni tramite modelli computazionali della teoria della mente. L’architettura del sistema si avvale di reti neurali, strumenti particolarmente efficaci nel riconoscimento di pattern visivi. Questo studio apre nuove prospettive per supportare i bambini con ASD nell’acquisizione di competenze sociali in contesti quotidiani. Tuttavia, la sfida principale è garantire che i metodi proposti siano in grado di adattarsi dinamicamente alle esigenze individuali, con un’attenzione particolare agli aspetti etici e alla sostenibilità delle soluzioni adottate.
Robot IA per l’autismo: Focus sul riconoscimento di schemi sociali e comprensione delle intenzioni.
Monica Di Domenico
;Alessia Sozio;Lucia Campitiello;Pio Alfredo Di Tore
2026
Abstract
This study extends the framework proposed by Campitiello et al. (2024) on the educational robot with AI, focusing on emotion recognition and social skills in children with autism spectrum disorder (ASD). The research explores additional key aspects according to Howlin et al.’s (1999) socio-cognitive development program. Specifically, it addresses the recognition of schematic designs as social icons and simplified facial expressions, as well as the understanding of desires and intentions through computational models of Theory of Mind. The system architecture utilizes neural networks, tools particularly effective in visual pattern recognition. This study opens new perspectives for supporting children with ASD in acquiring social skills in everyday contexts. However, the main challenge lies in ensuring that the proposed methods can dynamically adapt to individual needs, with particular attention to ethical considerations and the sustainability of the adopted solutions.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.


