A Social Robot-facilitated Performance Assessment of Self-care Skills for People with Alzheimer's: A Preliminary Study

Worldwide, there are approximately 10 million new cases of dementia reported each year. Due to impairments in cognitive functioning such as loss of memory, linguistic abilities, and problem-solving skills, persons living with dementia (PLWDs) may face challenges in accomplishing activities of daily living (ADLs). Socially assistive robotics (SAR) holds promise as an effective tool to assist PLWDs with their ADLs; however, developing a powerful, reliable SAR requires a robust dataset of PLWDs performing ADLs to ensure the efficacy and reliability of the SAR. In this paper, we present the development of a teleoperated robotic platform using a humanoid social robot. This platform provided instructions to perform ten Performance Assessment of Self-care Skills (PASS) tasks, while multimodal sensors were employed to collect the individual's physiological signals, behaviors, and interactions with the robot. We conducted a preliminary study to investigate the acceptability, user experience, and feasibility of the platform in a simulated home environment with five young cognitively normal individuals. Participants' experiences and satisfaction with the platform were evaluated through a questionnaire. The results demonstrate the capability of the robot administering PASS tasks. However, one limitation of the current platform is the low efficiency of controlling the robot to move in the Smart Home. Our next step is to implement navigational abilities in the robot and then conduct the experiments with a large cohort of PLWDs to generate our multimodal dataset. This work aims to contribute significantly towards the development of a powerful, reliable, and trustworthy SAR to help PLWDs with their ADLs.