Generative Responsiveness in Assistive Care Ecosystem - GRACE

The growing aging population presents significant challenges for delivering personalized, adaptive, and sustainable elderly care. Traditional care models often fall short in addressing the complex and dynamic needs of seniors, thus underscoring the need for integrated ecosystems that leverage both technological advancements and human capacities. This work introduces Generative Responsiveness in Assistive Care Ecosystem (GRACE), a human-AI collaborative framework currently under development. GRACE aims to support elderly individuals through continuous health monitoring, predictive analytics, emotional support, and dynamic service composition and evolution. Building on prior research, including the Elderly Care Ecosystem (ECE) concept and methodologies for service composition (SCoPE) and evolution (SEvol), GRACE is designed to facilitate the co-creation of evolving care solutions. The approach aims to integrate human stakeholders - seniors, families, caregivers, and healthcare providers - with AI agents in a hybrid network architecture. The framework comprises three distinct layers - Sensing and Acting, Cognitive and Analytics, and Interaction - fostering synergistic collaboration between human actors and AI. This paper outlines GRACE’s conceptual foundations, presents its initial functional design, and illustrates human-AI collaboration scenarios. It also discusses anticipated organizational, technological, and ethical challenges in real-world deployment and concludes by identifying key research directions for scaling, refining, and assessing the long-term impact of such AI-driven collaborative ecosystems in elderly care.