아두이노 기반 학습 과정의 오류 탐지를 위한 인공지능 모니터링 시스템 개발 = Development of an AI-Based Monitoring System for Error Detection in Arduino Circuit Learning

다국어 초록 (Multilingual Abstract)

This paper presents the development of an image-based artificial intelligence monitoring system designed to detect common wiring errors that occur in introductory Arduino circuit learning activities. The developed system analyzes circuit images captured by learners and classifies typical errors such as missing power connections, incorrect LED polarity, and misaligned jumper wires using a lightweight convolutional neural network (CNN). To improve interpretability, the system incorporates Grad-CAM visualization to highlight image regions that contribute to the model’s error classification. This visual feedback allows learners to understand not only whether an error has occurred but also where the error is located within the circuit, thereby supporting more effective self-correction during hands-on activities. The monitoring system was implemented and applied in an actual classroom environment where students constructed basic LED–resistor circuits. Experimental results indicated improvements in error recognition speed, self-correction ability, and conceptual understanding. Instructors also reported a reduction in repetitive debugging tasks and smoother class operation due to the system’s real-time visual feedback. These results demonstrate that the developed AI-based monitoring system can effectively support physical computing education by providing immediate and explainable feedback, highlighting its practical applicability in educational settings.

목차 (Table of Contents)

• Ⅰ. 서론 1
• 1.1 연구 배경 및 필요성 1
• 1.1.1 피지컬 컴퓨팅과 아두이노 교육 2
• 1.1.2 회로 오류의 시각적 특성 4
• 1.1.3 딥러닝 기반 회로 이미지 분석 5
• 1.1.4 Grad-CAM 기반 오류 위치 시각화 6
• 1.2 연구 목적 및 범위 7
• Ⅱ. 시스템 설계 및 구현 8
• 2.1 시스템 전체 구조 설계 8
• 2.2 이미지 획득 및 전처리 9
• 2.3 오류 분석 모듈 10
• 2.4 오류 시각화 모듈 11
• 2.5 사용자 인터페이스 및 동작 흐름 13
• 2.5.1 전체 동작 흐름 개요 13
• 2.5.2 사용자 인터페이스 14
• Ⅲ. 실험 결과 및 분석 15
• 3.1 실험 환경 및 과정 15
• 3.2 오류 발생 시스템 탐지 결과 16
• 3.3 단계별 오류 탐지 및 시각화 결과 17
• 3.4 교육 현장에서의 효과 18
• 3.4.1 학습자 측면 효과 18
• 3.4.2 교사 측면 효과 19
• Ⅳ. 결론 20
• 참고문헌 22
• ABSTACT 24