CAREER: Contextual Human-Congruent AI interactions: Crafting Shared Mind-Machine Representation through Cognitive Behavioral Architectures in Nursing

For robots to assist workers effectively in fields such as healthcare, they must perceive the environment in ways that align with human perspectives. Currently, artificial intelligence (AI) and robotic systems often fail to meet expectations when moving from laboratory settings to real-world applications. This results in frustration and inefficient use of resources. Workers have unique expertise in their tasks, making their input essential for developing supportive AI and robotic systems. This Faculty Early Career Development Program (CAREER) award seeks to advance robotic intelligence and intuitive human-robot collaboration by integrating worker insights. This will enable robots to interpret the world in ways consistent with human reasoning about real-world work. Additionally, the project aims to empower nurses, the largest group of healthcare professionals in the United States, by equipping them with tools to become technological innovators who can transform healthcare to improve national health and societal outcomes. The goal of this CAREER award is to achieve human-congruent embodied AI that uses human cognitive and behavioral insights to produce a shared mind-machine representation of complex, real-world work. This effort will involve a mixed-methods research approach for rich, holistic insight, using both lab and field studies including observations, co-creation, and human-robot interaction experiments. Research insights will lead to the creation and iterative refinement of a Cognitive-Behavioral Architecture (CBA) which incorporates the foundational cognitive architecture, ACT-R. Other components of the CBA include empirically-derived ethograms (i.e., systematic descriptions of observable behavior), along with common Machine Learning and statistical methods like machine vision and Bayesian inference to aid with Machine Theory of Mind and guide intuitive robot behaviors. The CBA will help robots to perceive real-world work in a way that is congruent with how human collaborators sense and act. This work has the potential to revolutionize how intelligent engineered systems can successfully collaborate with humans through shared mind-machine representations of complex work. 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.