The aging population faces growing challenges with Mild Cognitive Impairment (MCI), affecting 15–20% of adults over 65, with higher prevalence in lower-income areas, where access to tailored home modifications is limited, hindering kitchen navigation. This solution enables easy self-modification or DIY rearranging of homes for older adults with MCI and their care partners, addressing a critical need. Online kitchen images often lack MCI-specific features, prompting this study to adapt real kitchen images with Home Design Guidelines (HDG) aspects into AI-generated, accessible designs. We start with kitchens due to their vital role in activities of daily living (ADL) and instrumental activities of daily living (IADL) for this population. Using stable diffusion models with personalized image generation, we trained on 100 real kitchen images enhanced with HDG prompts to create layouts with open spaces, clear cabinetry, and non-slip flooring. We evaluated image accuracy using machine learning methods, including the Contrastive Language-Image Pretraining (CLIP) score (0.68-0.75) for semantic alignment and the Generated Image Quality Assessment (GIQa) score (0.45-0.65) for visual realism, indicating moderate to high alignment with accessibility goals. An online survey with 20-30 older adults with MCI and caregivers will deliver insights on usability, safety, and transformation potential using 4-6 randomized image pairs (standard vs. AI-modified), validating designs for converting existing kitchens. Future contributions include refining the model for broader adoption, expanding to other living spaces, and enhancing quality of life by providing accessible cognitive-impairment-friendly designs, aligning with AGen 2026 themes of Technology, Built Environment, and Resilience.

Authors:
Ibrahim Bilau, Georgia Institute of Technology, United States
Nicole Li, Georgia Institute of Technology, United States
Terrence Malayvong, Georgia Institute of Technology, United States
Eunhwa Yang, Georgia Institute of Technology, United States

About the Presenter(s)
Ibrahim Bilau is a GRA at Georgia Tech and Ph.D. candidate in Building Construction. His work explores technology and the built environment, focusing on HCI and cognitive aging, with current projects on MCI-friendly kitchen design.

Connect on Linkedin
https://www.linkedin.com/in/ibrahim-bilau/

See this presentation on the full schedule – On Demand Schedule

Bookmark