TRANSFORMING ELECTRONICS EDUCATION WITH AI: A CHALLENGE-BASED LEARNING APPROACH

The integration of Artificial Intelligence (AI) into electronics and computer science education offers new opportunities to enhance student engagement and learning outcomes.This paper presents a comprehensive approach to embedding AI-driven tools in laboratory-based courses, making complex concepts more tangible through hands-on experimentation.Developed by the GRIDS educational innovation group, this initiative leverages challenge-based learning and emerging technologies to create an interactive and adaptive learning environment.Three key AI applications serve as the foundation of this approach: Facial Emotion Recognition (FER), Human Activity Recognition (HAR), and Embedded Speech Technologies.These domains provide students with an opportunity to explore AI beyond theoretical concepts, engaging directly with machine learning models and real-world datasets.The Facial Emotion Recognition module enables students to experiment with deep learning for affective computing, using Convolutional Neural Networks (CNNs) and facial landmark detection to analyze emotions from video feeds.Through platforms like Google Colab and Raspberry Pi, students can train models with existing datasets (e.g., RAVDESS) or collect their own, gaining practical experience in AI model deployment.The Human Activity Recognition component introduces AI-powered gestures and movement analysis.By utilizing both inertial sensor data and image-based hand pose recognition, students implement deep learning architectures, such as Long Short-Term Memory (LSTM) networks and CNNs.This hands-on approach enhances their understanding of real-time AI applications in fields like robotics, healthcare, and human-computer interaction.The Embedded Speech Technologies module focuses on voice-driven interfaces, employing the Picovoice platform for speech recognition and intent detection.Using tools such as Porcupine Wake Word and Rhino Speech-to-Intent, students develop low-cost, voice-controlled systems that showcase the potential of AI for embedded applications.Across these projects, students engage in design, implementation, and optimization of AI models within resource-constrained environments, reinforcing problem-solving and critical thinking skills.The combination of service learning, design thinking, and adaptive learning methodologies, supported by platforms like Moodle, YouTube, and GitHub, ensures a scalable and accessible experience.Qualitative feedback highlights increased student motivation, deeper conceptual understanding, and stronger engagement with AI-based tools.The findings suggest that incorporating AI into laboratory settings not only enhances learning outcomes but also better prepares students for the evolving demands of AI-driven industries.