The UNESCO “Education 2030 Agenda” highlights the importance of inclusive, equitable, and personalized learning. In alignment with this vision, the Central Committee of the Communist Party of China released China’s Education Modernization 2035, which explicitly advocates for “advancing education modernization through informatization” and underscores the right of children and adolescents with disabilities to receive appropriate education.

In China, the prevalence of autism spectrum disorder (ASD) among children has reached nearly 2% and continues to rise rapidly. More than ten million individuals are affected, the majority of whom are unable to live independently. Autism has thus emerged as a pressing social issue. Marked by profound challenges in social interaction and communication, ASD currently has no effective medical cure and often leads to lifelong disability. As a result, education-based interventions remain the primary and most effective means of support. Meeting the educational needs of children with autism has become a key strategic priority in China’s broader education development agenda.

Compared with developed countries, China’s efforts in autism rehabilitation began relatively late. A shortage of qualified professionals has limited service capacity to fewer than 300,000 individuals. The labor-intensive nature of existing interventions further contributes to high costs—averaging as much as 70,000 yuan per person per year. However, advances in artificial intelligence are now creating new opportunities for innovation in autism education and intervention. While early-stage research in developed countries has shown the potential benefits of intelligent educational approaches, current practices remain rudimentary—often characterized by basic personalization and superficial evaluation. These methods tend to overlook the significant individual differences among autistic children, and fail to bridge the gap between fragmented, inefficient learning resources, rigid intervention models, and the complex, varied developmental needs of these learners.

Three persistent technological challenges continue to hinder progress in autism education interventions: enabling personalized expression, constructing adaptive resource systems, and improving intervention processes:

• Personalized Expression: Children with autism display wide-ranging individual differences. Traditional learner modeling techniques fall short in capturing the dynamic evolution of their cognitive abilities and emotional states, leading to low accuracy in personalized learning support.

• Resource Construction: The developmental trajectories of autistic learners are highly complex, often requiring multiple stages and varied scenarios for social interaction learning. Current educational resources are scattered and inefficient, and lack the design frameworks needed to align with the cognitive characteristics of autistic learners. This frequently results in cognitive overload that is difficult to manage.

• Process Intervention: Social interaction impairments in autism span multiple domains—behavioral, linguistic, and emotional—and involve a complex system of indicators. Existing intervention methods are often rigid and unsuitable, while commonly used assessment tools such as questionnaires and observational data lack objectivity and precision.

In response to these challenges, the National Engineering Research Center, together with the School of Computer Science at Wuhan University, Zhongnan Hospital of Wuhan University, and Qihui Special Education School in Wuhan, has launched a collaborative initiative to develop data-driven technologies, platforms, and applications for autism education intervention. Supported by a range of national and provincial-level programs—including the National Science and Technology Support Program, the National Key R&D Program, the National Natural Science Foundation of China, the National Social Science Fund, the Ministry of Education’s Humanities and Social Sciences Fund, and Hubei Province’s Major Technology Innovation Projects—this initiative focuses on leveraging data sensing and behavioral analytics to overcome key technological bottlenecks. These include modeling and interpreting the cognitive and psychological states of autistic learners, building structured human-computer interactive learning environments, and developing adaptive, personalized educational interventions tailored to individual needs.