Comparison of AI-Driven and AR-Driven Design for Disassembly (DfD) of a Shekili-Inspired Emergency Shelter Design

The research compares the application of Artificial Intelligence (AI, henceforth) and Augmented Reality (AR, henceforth) in developing a Shekili-inspired emergency shelter design. Shekili, with no glue or nail structure, can be adapted as an emergency shelter in earthquake and post-disaster situations. This paper aims to test the applicability of AI or AR in emergencies, considering that time and budget are pivotal aspects of consideration that should be addressed immediately. The research questions discussed in this paper are twofold. Firstly, the extent of AI and AR systems to facilitate DfD (Design for Disassembly, henceforth) in disaster relief. Secondly, the ease of use and learning from the perspective of users. The authors try to address the presented research gaps regarding the extent to which AI or AR may fill in times of emergency through shelter design. One potential solution which can be achieved is a designed structure that may bring modularity to the DfD principles. The findings illustrate that AR provides more detailed views of the elements than AI, as AI is based on images and textual information. On the other hand, AR is more thorough regarding the information provided. Future studies will include testing the AR system in real-scale emergency shelter designs and developing a guide for disaster relief assistance using a combination of AI and AR. Lastly, pedagogical benefits both for architectural students and laypersons are laid out.