Transforming Engineering Education Using Generative AI and Digital Twin Technologies
Project Overview
The document examines the integration of generative AI and digital twin technology in engineering education, emphasizing their potential to enhance learning experiences by aligning digital twin fidelities with Bloom's taxonomy. It highlights the advantages of employing low, medium, and high-fidelity digital twins tailored to different educational levels, which facilitate a more engaging and practical learning environment. Additionally, it discusses the role of large language models in offering personalized learning support, thereby catering to individual student needs and improving overall educational outcomes. The proposed framework aims to effectively connect theoretical knowledge with practical applications, ultimately enhancing the effectiveness of industrial education and preparing students for real-world challenges.
Key Applications
Digital Twin Technology and AI-based Virtual Tutor
Context: Undergraduate to doctoral level education, focusing on digital twin technologies for Industry 4.0, intelligent manufacturing, and personalized learning across various educational settings.
Implementation: Digital twin technologies ranging from low-fidelity to high-fidelity models are integrated with generative AI-based virtual tutoring systems. Low-fidelity digital twins introduce basic concepts, medium-fidelity models engage in production planning, and high-fidelity models support advanced research with real-time data integration. The AI tutor provides personalized guidance and feedback across all educational levels.
Outcomes: Students develop foundational knowledge in industrial systems, gain practical experience in real-world scenarios, and enhance their analytical and problem-solving skills. The AI tutor improves learner engagement and supports differentiated instruction tailored to individual needs.
Challenges: Limited realism in low-fidelity models may not fully simulate actual system behavior; complexity in creating medium-fidelity simulations; high data costs and the need for continuous adaptation of high-fidelity models; ensuring the AI tutor provides accurate and relevant real-time support.
Implementation Barriers
Technological Barrier
Difficulty in accurately simulating real-world systems due to the limitations of low and medium-fidelity digital twins. The complexity involved in integrating digital twin technology with existing educational frameworks.
Proposed Solutions: Advancements in machine learning and computational power can improve the accuracy and adaptability of digital twins. A structured approach using frameworks like Bloom's taxonomy can guide integration efforts.
Cost Barrier
High costs associated with developing and maintaining high-fidelity digital twins.
Proposed Solutions: Leveraging generative AI to reduce data requirements and enhance predictive capabilities can mitigate costs.
Project Team
Yu-Zheng Lin
Researcher
Ahmed Hussain J Alhamadah
Researcher
Matthew William Redondo
Researcher
Karan Himanshu Patel
Researcher
Sujan Ghimire
Researcher
Banafsheh Saber Latibari
Researcher
Soheil Salehi
Researcher
Pratik Satam
Researcher
Contact Information
For information about the paper, please contact the authors.
Authors: Yu-Zheng Lin, Ahmed Hussain J Alhamadah, Matthew William Redondo, Karan Himanshu Patel, Sujan Ghimire, Banafsheh Saber Latibari, Soheil Salehi, Pratik Satam
Source Publication: View Original PaperLink opens in a new window
Project Contact: Dr. Jianhua Yang
LLM Model Version: gpt-4o-mini-2024-07-18
Analysis Provider: Openai