Haru: The Social Robot
A Honda Research Institute UX Research Design in Human-Robot Interaction & Children's Education
ROLE
Human-AI Interaction UX Researcher
TOOLS
Mural, Nvivo, Microsoft 365,
Indesign & llustrator
TIMELINE
10 weeks
TIMELINE
3 Researchers, 1 Psychologist, 4 Engineers
PROJECT OVERVIEW
In Context
1. WHAT
User Experience Research Design to investigate American kids and teachers experiences with Haru in the classroom and if this interaction has any impact on improving communication skills on school age kids.
2. USERS
Elementary and Middle School
Kids and Educators from the United States
3. CONSTRAINS
The robot's delayed arrival postponed school-based research. We conducted internal simulations to refine activities and solidify the research framework ahead of fieldwork.
4. OUTCOMES
-
Designed child-centred research tools (assessments & focus groups), refined via pilot tests & teachers' feedback.
-
Designed a multimodal analysis framework for consistent, reliable analysis outcomes.
5. IMPACT
User Experience Research Design to investigate American kids and teachers experiences with Haru in the classroom and if this interaction has any impact on improving communication skills on school age kids.
6. OWNERSHIP
-
Led secondary research on best practices for child-centred surveys
-
Designed co-design workshops and focus groups
-
Developed assessment workbooks and an Induction Coding Map
-
Built a thematic analysis workflow for multiple coders
THE PROJECT
Meet Haru
Haru is a social robot, designed to foster emotional connections with humans. It has been designed, developed and intensely researched over the past 7 years by Honda Research Institute (HRI) Japan, through a multidisciplinary collaboration of engineers, designers, academics, researchers, artists, and animators all over the world.
In 2025, a new phase of the Haru Project started, and HRI USA was invited to join in contributing to AI-Human interaction user experience research with teachers and kids from elementary and middle school. For this new phase, the goal was to investigate how Haru could foster the improvement of communication skills in schools.
The Best of Our Features
This is the space to introduce the Features section. Use this space to highlight your unique aspects and to present specific credentials, benefits or special features you offer.
Feature Title
Share your feature information here to attract new clients. Provide a brief summary to help visitors understand the context and background, and add details about what makes this feature significant.
Feature Title
Share your feature information here to attract new clients. Provide a brief summary to help visitors understand the context and background, and add details about what makes this feature significant.
Feature Title
Share your feature information here to attract new clients. Provide a brief summary to help visitors understand the context and background, and add details about what makes this feature significant.
Feature Title
Share your feature information here to attract new clients. Provide a brief summary to help visitors understand the context and background, and add details about what makes this feature significant.
THE CHALLENGE
Communication Issues
The communication skills of school-age children have been impacted by the recent global shutdown due to Covid.
KEY CHALLENGES
01
Limited opportunities for practicing.
02
Need for constant guidance and feedback
03
Difficulty Expressing Oneself
04
Shyness and Social Inhibition
05
Learned Helplessness
How Might we
Design human-robot interactions between local groups in unstructured pedagogical settings to address target children’s communication skills?
This question guided the entire research and design process across both phases of the Haru Project.
OUR APPROACH
Research Plan
June - Aug 2025
Literature review on social robots and children, K-12 trends in education, pedagogical approaches.
Co-design workshop with Elementary and Middle School teachers.
Aug - Sep 2025
+ I WORKED IN THIS PHASE
Exploration of analysis approaches for the Haru/students/teacher sessions.
Design and first pilots of Haru sessions.
Ideation of the documentation and analysis plan
Secondary Research on reliability in intercoder analysis, surveying kids, social robots, K-12 group activities.
2026
Execution and documentation of the Haru sessions.
Data analysis.
I joined the project during its second phase for ten weeks, supporting focus group session design, shaping the analysis plan, and conducting secondary research to inform decisions.
CHILD-CENTRED DESIGNED
Research Tools
For this study, we used a mixed-methods design to collect qualitative and quantitative data during focus groups with children, the robot, and the teacher in a controlled environment. We designed self-report assessments and planned video and audio recordings of interactions with Haru through the activities we created for the focus groups, as well as field notes.This session highlights the process of designing the child-centred self-report assessments and the ideation of focus group activities.
1.0 Child-Centred Self-Report Tool Design
Context
The need for an assessment workbook to complement observational tools during workshops, capturing children's authentic experiences in a structured, playful, and engaging way.
Challenge
-
Designing reliable, age-appropriate surveys is complex.
-
Younger participants may misinterpret or answer to please adults.
-
Traditional Likert scales are often misused or biased.
Insights
-
Children provide feedback that adults cannot.
-
Surveys should be short, clear, and age-appropriate.
-
A 5-level smiley scale reduces response bias.
-
Embodied surveys support playful, reflective feedback.
(Sylla et al., 2017; Hall et al., 2016; Bell, 2007).
1.1 WORKBOOKS
Based on previous experiences working with children and on secondary research, I designed a child-centred assessment workbook to replace traditional surveys and gather meaningful insights.
Here you can see three examples of activities presented in the workbooks.
ACTIVITY 1
My blob-self
An irregular shape children fill with words or drawings about their feelings, interests, and dreams. It gives kids a low-pressure space to self-express while generating genuine insights into their inner world.
-
ACTIVITY 2
5 Levels of Happiness
A 5 levels of happiness scale where children rate themselves across six communication aspects. Reduces response bias common in standard scales by making self-evaluation age-appropriate and visual.
Bell, A. (2007). Designing and testing questionnaires for children. Journal of Research in Nursing, 12(5), 461–469.
ACTIVITY 3
The Paper Ladder
A ranking tool where children place picture post-its on ladder steps from liked least to liked most. Makes evaluation playful and hands-on while capturing authentic preferences and thought processes.
Sylla, C. M., Arif, A. S., Segura, E. M., & Brooks, E. I. (2017). Paper ladder: A rating scale to collect children's opinion in user studies. In Proceedings of the 19th International Conference on Human-Computer Interaction with Mobile Devices and Services.
IMPACT
I redesigned traditional survey instruments into child-centred multimodal workbooks to increase engagement, reduce interpretive bias, and strengthen the validity of qualitative and quantitative responses.
DESIGNED FOR KIDS
Research Tools
Research Plan
For this study, we used a mixed-methods design to collect qualitative and quantitative data during focus groups with children, the robot, and the teacher in a controlled environment. We designed self-report assessments and planned video and audio recordings of interactions with Haru through the activities we created for the focus groups, as well as field notes.
This session highlights the process of designing the child-centred self-report assessments and the ideation of focus group activities.
2.0 Focus Groups Activities Ideation
01
Grounding
Working with a teammate and drawing from secondary research and co-design workshops with local teachers, we designed age-specific activities for elementary and middle school, making sure each one addressed real communication challenges between the robot, teachers, and students.
02
Systematization
We built a structured Mural matrix to classify and assess each activity across features, robot roles, and ecological feasibility, turning a large set of ideas into a clear, research-informed structure
IMPACT
The matrix made prioritization transparent, supported team voting and teacher feedback, and gave us a solid foundation to turn raw ideas into a defensible research plan.
THINKING OF ANALYSIS
Analysis Framework
A structured approach to turning 24 hours of video, workbooks, and transcripts into consistent, defensible insights across a distributed research team.
01
Data & Rationale
6 schools (3 elementary, 3 middle), 3 sessions each (18 total). One continuous group per school: 3–4 students + 1 teacher. Data: quantitative and qualitative, from workbooks, transcripts, and ~24 hrs of 360° videos from focus group sessions.
02
Code Map
We designed an initial deductive coding list to capture emotional affect, participant roles, engagement, turn-taking, and conflict markers. I built a code map to visualise connections, define code rules, and support real-time refinement during analysis.
03
Automation
The team explored automated transcription and sentiment extraction across text and audio. Outputs are treated as probabilistic signals — not conclusions — to support, not replace, human interpretation.
04
NVivo
All materials were organised in NVivo to enable thematic synthesis and consistent analysis across the team. I guided researchers through the process as we built the coding structure together.
.png)
IMPACT
By visualising the coding system, we could quickly spot gaps and refine codes in real time, targeting a comprehensive codebook for the future.
THINKING OF ANALYSIS
Designing Trust in Qualitative Analysis
During a test, intercoder agreement reached 70%, revealing ambiguity in the deductive coding system.
01
Gap Identified
The team was exploring automation to support analysis, but neither the human coding system nor the automation pipeline had been fully validated. In a child-centered qualitative study, interpretive reliability was essential.
02
My Role
conducted a focused literature review on intercoder reliability and qualitative best practices, identified strategies to reduce ambiguity, and designed a structured workflow grounded in research. I also documented decision rules and procedures to ensure clarity across coders.
03
Current Status
Current Status Data collection was delayed, so the workflow hasn't been fully tested yet — but the project now has a reliability-first framework ready for implementation.
