NSF Funds $21.5 Million AI Institute to Create 'Adaptive Teammates' for Healthcare
A five-university consortium will develop collaborative AI systems and digital twins to assist clinicians in personalized patient rehabilitation.
The National Science Foundation has awarded a $21.5 million grant to establish the AI Institute for Human-AI Cooperation (NSF HAIC). This five-year initiative, operating under Cooperative Agreement No. 2433450, is designed to integrate artificial intelligence as a collaborative partner in clinical settings, shifting the role of AI from a standalone tool to an "adaptive teammate" capable of supporting healthcare providers in personalized medicine and rehabilitation.
Headquartered at the University of Delaware’s Data Science Institute, the effort is led by the University of Delaware in partnership with the University of Pennsylvania, the University of Florida, Princeton University, and Delaware State University. The institute will bring together more than 40 partners spanning government, industry, healthcare, nonprofit organizations, and academia. A core technical component of the initiative involves the use of AI-enabled "digital twins"—virtual avatars that allow clinicians to test and refine treatment options in a simulated environment before applying them to real-world patients.
The Shift Toward Human-Centered AI
This initiative arrives as the medical field seeks to move beyond rigid, automated tools toward human-centered AI. The institute focuses specifically on physical therapy and rehabilitation, where patient needs are highly variable. Rather than replacing human decision-making, the system is designed to integrate multimodal data, including video feeds, clinical notes, and information from wearable sensors. This approach allows the AI to provide real-time, personalized support that adapts to the specific physical requirements of a patient's recovery process.
Implications for Clinical Practice
By treating AI as a teammate rather than a software utility, the institute aims to improve recovery outcomes for patients recovering from major surgeries or strokes. The focus on continual learning and trust ensures that the AI evolves alongside the clinician's expertise. The goal is to develop systems that act as "true clinical partners" in real-world settings. This transition is expected to not only enhance patient care but also train a new generation of healthcare professionals to work effectively alongside intelligent systems.
Future Outlook
As the five-year project progresses, the consortium will focus on the dynamic collaboration between human users and AI. The objective is to transform AI into a system that can interact in dynamic collaboration with users. Observers will be watching how these digital twins translate into measurable improvements in patient mobility and whether the multi-institutional model can be scaled to other medical specialties beyond physical rehabilitation.