Abstract—People with Dementia (PwD) face considerable cognitive challenges that hinder their ability to perform daily activities, commonly referred to as Activities of Daily Living (ADLs), and encounter even greater difficulties with the more complex Instrumental ADLs (IADLs) such as cooking and housekeeping. Without sufficient support, PwD risk losing their independence, making them more vulnerable to loss of autonomy. Conversely, as PwD experience increasing difficulty in understanding their circumstances, caregivers are relied on heavily for all aspects of support leading to caregiver burden. To tackle these issues, a range of Intelligent Assistive Technologies (IATs) have been developed to help PwD manage some IADLs independently. However, most current IATs offer limited support for IADLs due to the lack of understanding of the unique communication needs of PwD and insufficient user-specific customization. This paper presents an innovative homecare assistive device, CATcare (Cognitive Assistive Technology Care), specifically designed for people with mild dementia and their caregivers. The CATcare system, which integrates ChatGPT, location classification, and object detection, provides a context-based understanding of PwD’s needs. It is designed for easy operation and customization through smartphones or smart glasses, is easily tailored to meet specific IADL requirements, offers step-by-step guidance, prompts, and timely feedback to assist PwD in completing their daily tasks. Our research leverages transfer learning from recent Artificial Intelligence (AI) models for indoor localization, object detection, and large language models. The findings underscore the potential promise of developing a customizable and personalizable CATcare tool, aimed at improving the quality of life for PwD while easing the burden on caregivers.
 
Keywords—people with mild dementia, intelligent assistive technologies, instrumental activities of daily living, audio/visual prompting, transfer learning, ChatGPT, indoor localization


Copyright © 2025 by the authors. This is an open access article distributed under the Creative Commons Attribution License which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited (CC BY 4.0).