Integrating Traffic Prediction STEM Cases into Foundation Courses: A Tailored Approach for Information-Oriented Engineering Education

With the rise of AI and data-driven engineering, traditional foundational courses such as Engineering Mathematics often fall short in preparing students for interdisciplinary, real-world challenges. These courses tend to emphasize theoretical derivations with limited relevance to practical applications, resulting in disengaged learners and weak knowledge transfer. To bridge this gap, we designed and integrated an LSTM-based traffic flow prediction STEM module into the Engineering Mathematics curriculum at BUPT. The module links core mathematical topics to a hands-on machine learning application. To validate its effectiveness, a survey from students showed high levels of interest and perceived relevance, particularly in mathematical modeling and visualization. These results suggest that embedding domain-specific, AI-driven STEM cases can enhance student engagement, foster applied mathematical thinking, and promote interdisciplinary competence in undergraduate engineering education.