CRII: HCC: RUI: AI-Driven Collaborative Goal Setting with Cognitively Assistive Robots

The project involves application of emerging artificial intelligence (AI) technologies to engage in collaboration with a human. The collaborations involve adaptive goal setting, which is a new frontier in Human-AI Interaction. The project will develop a human-AI system to support individuals with mild cognitive impairments (MCI). At the core of the project are socially assistive robots designed to help patients set their goals in a collaborative way. This work aims to increase the efficacy of these interventions and ultimately improve people's emotional well-being, promote autonomy and self-efficacy, and better their quality of life. The main goal of this project is to develop a transformational goal-centered adaptation framework for autonomous behavior adaptation to support users with achieving real-world goals in longitudinal rehabilitation interventions. Our system will comprise three core properties. First, we will design a goal-setting framework using multimodal, naturalistic data for modeling each user's unique goal progress and motivation based on their diverse real-world behaviors and interactions with a robot. Second, we will develop human-centric approaches for automatically generating and recommending subgoals to support a person's overarching rehabilitation goals. Third, we will develop personalization methods to automatically adapt the behavior of a robot delivering a longitudinal behavioral health intervention to support a user's goal progress and motivation. Ultimately, these contributions will advance the field's understanding of how robots can conceptualize and adapt their behavior to people with cognitive impairments, grounded in current clinical practice. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.