2025 ASEE Annual Conference & Exposition

This paper aims to present the integration of FPGA-based AI acceleration into the education of predictive maintenance systems, emphasizing its relevance to engineering graphics, system design, and automation. The primary objective is to develop educational frameworks that allow students and professionals to better understand and implement AI-driven predictive maintenance using FPGAs. By equipping learners with practical experience in developing, deploying, and analyzing these systems, educational institutions can better prepare the future workforce to meet the demands of Industry 4.0. FPGAs enable rapid data processing and support graphical representation of machine health, anomaly detection, and predictive outcomes. Engineering students gain hands-on experience designing systems that require precise timing and data flow management, skills critical in manufacturing, mechanical, and electrical engineering. This study involves the development of educational modules that simulate real-world predictive maintenance scenarios. Assessment involves both quantitative metrics (speed, energy consumption) and qualitative feedback from students on learning outcomes. Initial findings indicate that integrating FPGAs significantly reduces latency in AI model execution, enabling faster and more accurate predictive maintenance decisions. Students who participated in the FPGA-based modules demonstrated a deeper understanding of system-level design and were able to visualize data flow and processing better than those using traditional processors. Additionally, the hands-on experience with FPGA hardware improved their problem-solving and design skills, making them better prepared for careers in industrial automation.