This project is actively developing a data-driven platform that integrates Digital Twins (DTs) and Large Language Models (LLMs) to assess exoskeletons in construction. Using Retrieval-Augmented Fine-Tuning (RAFT), a GPT-4o Mini model has been trained on peer-reviewed research and industry data. The goal is to create a user-friendly decision-making tool that enables construction stakeholders to evaluate exoskeleton performance based on human factors such as cognitive load, biomechanics, and ergonomics.
Akanmu, A., A., Afolabi, A., Okunola, A. (2023). Human-in-the-loop digital twin framework for assessing ergonomic implications of exoskeletons. International Conference on Construction Applications of Virtual Reality. https://doi.org/10.36253/979-12-215-0289-3.121.
Methodology
