ARtonomous: Introducing Middle School Students to Reinforcement Learning Through Virtual Robotics
Project Overview
The document discusses the transformative role of generative AI in education, highlighting innovative tools like ARtonomous, a virtual robotics education platform aimed at middle school students. This tablet-based application introduces learners to reinforcement learning (RL) and machine learning (ML) concepts, effectively reducing the financial barriers typically associated with traditional robotics education. By enabling students to train virtual robots and program their behaviors, ARtonomous not only enhances understanding of complex AI concepts but also significantly boosts student engagement and curiosity in the field of machine learning. The findings indicate that such tools can foster a deeper interest in STEM education while providing an accessible and interactive learning experience, ultimately addressing obstacles that have historically limited participation in robotics. Thus, the integration of generative AI in educational contexts is proving to be a promising avenue for enriching learning experiences and expanding student opportunities in advanced technology fields.
Key Applications
ARtonomous - a tablet-based application for training virtual autonomous robots using reinforcement learning.
Context: Middle school education, targeting students aged 11-14.
Implementation: Students use the ARtonomous app to create virtual robot models, train them using RL, and write code to control their behavior.
Outcomes: Students developed an understanding of reinforcement learning, showed high engagement levels, and expressed curiosity to learn more about machine learning.
Challenges: Initial knowledge gap regarding AI/ML concepts among students; reliance on general-purpose devices for training without specialized hardware.
Implementation Barriers
Cost Barrier
High costs of physical robotics kits and specialized hardware limit access for economically disadvantaged students.
Proposed Solutions: ARtonomous uses virtual robots that run on general-purpose devices, significantly reducing costs and increasing accessibility.
Knowledge Barrier
Students often lack understanding of AI and ML concepts, which can hinder engagement and learning.
Proposed Solutions: The ARtonomous application includes teaching materials that help build foundational knowledge of reinforcement learning.
Project Team
Griffin Dietz
Researcher
Jennifer King Chen
Researcher
Jazbo Beason
Researcher
Matthew Tarrow
Researcher
Adriana Hilliard
Researcher
R. Benjamin Shapiro
Researcher
Contact Information
For information about the paper, please contact the authors.
Authors: Griffin Dietz, Jennifer King Chen, Jazbo Beason, Matthew Tarrow, Adriana Hilliard, R. Benjamin Shapiro
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