Case ReportPublished empirical studyApr. 26, 2025
AAB-CASE-2025-RV-044
BrickSmart: Leveraging Generative AI to Support Children’s Spatial Language Learning in Family Block Play
Tsinghua Future Lab et al.; CHI 2025.
This page documents an AI literacy or AI education case for registry purposes. It is descriptive and does not imply AAB endorsement of any specific tool, provider, or intervention.
01
Implementation
University labs (Tsinghua Future Lab + partners)
02
Learning context
Informal learning
03
AI role
Tutor
04
Outcome signal
Spatial language
Registry Facets
0
Education Level
- Pre-K
- K-5
Subject Area
- Early math / spatial reasoning
- GenAI
Use Case Type
- Family learning tool
Stakeholder Group
- Parents
- Students
AI Capability Type
- Generative AI
Implementation Model
- Informal learning
Evidence Type
- Comparative study
Outcomes Domain
- Spatial language
- Engagement
Implementing Organization
1
Organization Type
University labs (Tsinghua Future Lab + partners)
Location
China / international co-authors
Primary Facilitator Role
Researchers
Learning Context
2
Setting Type
- Informal learning
Session Format
Family sessions with BrickSmart vs control
Duration
Comparative user study
Group Size
12 parent–child pairs reported in abstract
Devices
GenAI + block play + 3D models
Constraints
- Small n
- Language/cultural context
- GenAI dependence risks
Learner Profile
3
Age Range
Young children in block-play band (per study)
Prior AI Exposure Assumed
Low–mixed
Prior Programming Background Assumed
None
Educational Intent
4
Primary Learning Goals
- Improve spatial language via guided block play
- Support parents lacking expertise
- Structure interaction in three steps
Secondary Learning Goals
- Leverage GenAI responsibly for early cognition
What This Was Not
- Not school-wide RCT
AI Tool Description
5
Tool Type
BrickSmart GenAI coaching for block building
AI Role
- Tutor
- Co-creator
Languages
Study context (China-based team)
User Interaction Model
- Personalized instructions
- Vocabulary prompts
- Progress tracking
Safeguards
- Child-appropriate outputs
- Parent oversight
- Data minimization for child images if any
Activity Design
6
Activity Flow
- Three-step workflow
- Compare conditions
- Measure spatial language gains
Human Vs AI Responsibilities
- Parents remain play partners; AI scaffolds language
Scaffolding Strategies
- Structured phases reduce parental burden
Observed Challenges
7
Educators Reported
- Parents lack spatial pedagogy expertise
- GenAI quality variability
- Scaling hardware/toy requirements
Design Adaptations
8
Adaptations
- GenAI embedded in classic developmental toy play
Reported Outcomes
9
Engagement
- Designed for joint engagement
Learning Signals
- Comparative gains reported for spatial language
Educators Reflection
Model for responsible GenAI in family STEAM.
Ethical & Privacy Considerations
10
Privacy
- Child safety and content filters
- Commercial GenAI ToS for minors
- Equity of toy access
Evidence Type
11
Evidence
- Post assessment
- Activity documentation
- Practitioner observation
Relevance to Research
12
Potential Research Use
- Larger diverse family trials
- Teacher-mediated classroom variant
Relevant Research Domains
- Family learning
- Spatial reasoning
- GenAI
Case Status
13
Case Status
- Completed
AAB Classification Tags
14
Age
Early elementary
Setting
Home
AI Function
Language + spatial scaffolding
Pedagogy
Guided play
Risk Level
Medium
Data Sensitivity
Medium
Registry Metadata
15
Case ID
AAB-CASE-2025-RV-044
Publication Status
Published empirical study
Tags
casePre-KChina / international co-authorsInformal learningGenerative AIEarly math / spatial reasoningGenAIFamily learning tool