The abstract should summarize the contents of the paper in short terms, i.e. 15–250 words. In recent years, computer architecture instruction has increasingly adopted project-based approaches to bridge the gap between theory and practical applications. This paper presents two hands-on learning activities using the Raspberry Pi platform to introduce students to real-time object detection and GPIO-based hardware control. These projects emphasize the integration of software and hardware, aligning with key concepts in embedded systems and edge artificial intelligence (AI). Students implement live video processing and physical feedback using lightweight tools such as Python, OpenCV, and YOLOv8. These projects, embedded in a computer architecture course, help students apply architectural concepts while tackling real-world challenges in deploying AI on resource-constrained devices. Informal feedback and classroom observations suggest increased engagement, improved confidence in embedded programming, and a deeper understanding of software-hardware interaction. This work offers a scalable, low-cost model for integrating embedded and AI topics into engineering curricula.