BEYOND THE BOOKS: A STUDY ON AI AUGMENTATION IN CIVIL ENGINEERING PEDAGOGY FOR PAVEMENT DESIGN

BEYOND THE BOOKS: A STUDY ON AI AUGMENTATION IN CIVIL ENGINEERING PEDAGOGY FOR PAVEMENT DESIGN
Javier Vasquez-Monteros, Universidad Técnica Particular De Loja, Ecuador / Yasmany García-Ramírez, Universidad Técnica Particular De Loja, Ecuador

(Presented at the International Conference on Research in Education and Science (ICRES) which took place on April 27-30, 2024, in Antalya, TURKEY (https://www.2024.icres.net/) and at the International Conference on Education in Mathematics, Science and Technology (ICEMST) (https://www.2024.icemst.com/) organized by the International Society for Technology, Education and Science (ISTES) http://www.istes.org).

This study delves into the transformative landscape of artificial intelligence (AI) with a focus on the integration of ChatGPT 3.5 and BARD in civil engineering education, specifically within the pavement design domain. The research addresses the pivotal question of whether AI tools can enhance the learning experience for students in this specialized field. Conducted with two cohorts of seventy undergraduate students enrolled in the Structural Pavement Design course at UTPL during the academic periods of October 2023 and February 2024, the methodology involved generating prompts covering various pavement design topics. Students evaluated AI-generated responses using a predefined grading scale based on provided literature, resulting in scores of 1 for correct, 0.5 for vague, and 0 for incorrect answers. Results indicated comparable responses from both ChatGPT 3.5 and BARD, with the latter receiving slightly higher overall ratings. Notably, students expressed that the use of AI tools significantly enhanced their understanding of course material. Interestingly, if considered as students, both ChatGPT 3.5 and BARD would have successfully passed the course based on their collective performance. Conclusions underscore the positive impact of incorporating AI tools into education, particularly in the realm of pavement design. While students reported an enhanced learning experience, the study acknowledges the existing limitations of AI technologies. The findings emphasize the need for a nuanced approach, recognizing potential benefits while mindful of associated constraints. In summary, this research asserts that integrating ChatGPT 3.5 and BARD as educational supplements positively influences the learning outcomes of civil engineering students specializing in pavement design. The study contributes valuable insights into the evolving role of AI in education, opening avenues for future exploration and refinement of these technologies within the academic domain.