01
Experienced Youth Mentors
Young volunteers who know how to communicate with autistic children and keep sessions warm, simple, and responsive.
Project Intro
Autism AI Social Skills Training Media System
Built by people who understand autism firsthand, combining technology, care, and lived experience to help children practice expression, perception, and connection.
This page presents the project in English as a standalone reference for visitors, partners, and stakeholders who need a clear overview of the team, the problem, the solution, the target users, and the delivery plan.
Who We Are
We are a team made up of youth volunteers, caring parents, autism families, and AI researchers and developers. Some of us have spent years helping autistic children assemble robots by hand. We know what they like, what they fear, what rhythm helps them relax, and what kind of feedback helps them keep moving forward.
Why Us
Four real capabilities make this project practical, not just aspirational.
02
Caring Parent Network
Parents who understand the daily reality of training, and who can provide grounded feedback and co-creation support.
03
Cooperative Family Community
Families willing to participate in validation, scenario testing, and iteration, which is hard to replicate in a lab.
04
Core AI R&D
Independent AI product development capability that combines language models with child-friendly training scenarios.
Pain Points
Autistic children face four core challenges in social skills training:
01
Limited Personalized Content
Existing stories and practice cards are often one-size-fits-all, so they cannot adapt to each child's interests or daily state.
02
Training Is Hard To Sustain
Without structured tools, training records become scattered and progress is hard to measure, which weakens consistency.
03
Emotion Recognition Is Hard
Emotion understanding is central to social interaction, but most tools are too abstract to support repeated low-stress practice.
04
Home-School Collaboration Breaks Down
Families and teachers often work in parallel instead of in sync, which reduces the impact of training across environments.
Solution
An AI-based media platform for social skills training across home and school settings.
1
AI Social Story Generation
Generate personalized social scenarios based on age, interests, and training goals, with image panels and voice reading support.
2
Emotion Recognition Training Module
Use structured practice and instant feedback to help children learn to identify and express emotions in a low-pressure environment.
3
Training Logs and Progress Reports
Let parents and teachers record observations, while the system summarizes weekly progress so improvement stays visible.
4
Content Safety Review
All AI-generated content is constrained by child-safe templates and a human review flow to protect sensitive users.
Audience
Parents
Core Need
A practical home training tool.
Platform Value
Generate personalized stories, save materials, and review weekly reports.
Teachers
Core Need
Lightweight classroom support.
Platform Value
Manage students, assign content, record observations, and export stage-based reports.
Administrators
Core Need
Content quality and safety control.
Platform Value
Review content, manage users, and monitor platform activity and compliance records.
Delivery Plan
An adaptive 6-12 month delivery model, guided by actual product progress.
M1
Definition and Design Freeze
Confirm user roles, page map, core flows, data model, and interface draft.
M2-3
Core Infrastructure Setup
Build child profiles, content structure, story generation flow, and initial admin tools.
M4-5
Main Workflow Launch
Ship story pages, result display, content library, training logs, and weekly summaries.
M6
Verifiable MVP Release
Complete QA, review queue, permission control, and bug fixes before pilot use.
M7+
Optimization and Expansion
Add more scenarios, richer reports, improved prompts, analytics, and organization features after validation.