Feasibility study for AI-Driven Decision Support System for Personalised Housing Adaptations and Assistive Technology

Abstract This feasibility study explores the development of the HomeAI-Enable system, an AI-driven decision support platform designed to provide personalised housing adaptations and assistive technology recommendations for aged and individuals with disabilities. The study follows a constructive research approach, in addition to development research techniques like focus groups and brainstorming exercises. Three different datasets comprising one, five, and over 500 diseases sourced from reputable databases such as the NHS and Mayo Clinic were used. The datasets include associated housing adaptation and assistive technologies for the symptoms and diseases. The system integrates these datasets to offer personalised solutions based on disease, body functions, symptoms, and environmental factors, using advanced techniques like fuzzy matching, natural language processing, and advanced machine learning models. The feasibility study involves the assessment of multiple AI models to determine their effectiveness in providing tailored recommendations for assistive technology and housing adaptations. A proof of concept was developed, and early tests suggest that the system aligns well with universal design principles and the needs of local authorities, rehabilitation specialists, and healthcare professionals, indicating the feasibility of implementing this solution in real-world environments. HomeAI-Enable will transform how needs are assessed, and solutions are delivered in housing adaptation through streamlining decision-making by the local authorities. It will provide occupational therapists with intelligent, data-backed recommendations that complement their assessments, enhancing confidence in housing adaptation choices.