Skip to main content Skip to navigation

Echo-Teddy: Preliminary Design and Development of Large Language Model-based Social Robot for Autistic Students

Project Overview

The document highlights the innovative use of generative AI in education through the introduction of Echo-Teddy, a Large Language Model (LLM)-based social robot designed specifically to aid autistic students in enhancing their social and communication skills. Unlike traditional scripted chatbots, Echo-Teddy employs both verbal and non-verbal cues for more adaptive interactions, which can significantly improve engagement and learning outcomes for these students. The design of the robot is rooted in key principles such as customizability and ethical considerations, ensuring that it meets the diverse needs of its users while maintaining a responsible approach to AI in education. Additionally, the implementation utilizes cost-effective hardware like Raspberry Pi to enhance scalability, making the technology accessible to a broader range of educational settings. The study underscores the transformative potential of LLMs in the realm of social robotics for special education, providing valuable insights for future developments in this field. Overall, the findings suggest that generative AI can play a crucial role in creating supportive learning environments, particularly for students with unique challenges, thereby fostering better educational experiences and outcomes.

Key Applications

Echo-Teddy, a social robot for autistic students

Context: Designed for use in educational settings, particularly for autistic children to enhance social interaction skills.

Implementation: Developed using Raspberry Pi with integrated LLM capabilities for context-aware interactions and combined verbal and non-verbal communication.

Outcomes: Improved engagement in social interactions, customized learning experiences, and potential scalability for broader use.

Challenges: High development costs, processing delays affecting interaction speed, and the need for ongoing evaluation of real-world effectiveness.

Implementation Barriers

Technical Barrier

Processing delays of 6-10 seconds during communication can disrupt engagement for autistic students.

Proposed Solutions: Upgrade hardware for faster local processing or split audio data for sequential speech synthesis.

Cost Barrier

High production and implementation costs limit accessibility and widespread adoption of social robots.

Proposed Solutions: Develop open-source versions of Echo-Teddy for DIY construction to allow customization and reduce costs.

Project Team

Unggi Lee

Researcher

Hansung Kim

Researcher

Juhong Eom

Researcher

Hyeonseo Jeong

Researcher

Seungyeon Lee

Researcher

Gyuri Byun

Researcher

Yunseo Lee

Researcher

Minji Kang

Researcher

Gospel Kim

Researcher

Jihoi Na

Researcher

Jewoong Moon

Researcher

Hyeoncheol Kim

Researcher

Contact Information

For information about the paper, please contact the authors.

Authors: Unggi Lee, Hansung Kim, Juhong Eom, Hyeonseo Jeong, Seungyeon Lee, Gyuri Byun, Yunseo Lee, Minji Kang, Gospel Kim, Jihoi Na, Jewoong Moon, Hyeoncheol Kim

Source Publication: View Original PaperLink opens in a new window

Project Contact: Dr. Jianhua Yang

LLM Model Version: gpt-4o-mini-2024-07-18

Analysis Provider: Openai

Let us know you agree to cookies