Collaborative Research: CyberTraining: Implementation: Small: Upskilling Future Workforce for Agentic Cyberinfrastructure-Enabled Manufacturing Ecosystem

The transition to Industry 4.0 has digitized modern manufacturing, yet the ability to autonomously interpret data and make dynamic decisions remains a critical bottleneck. Current manufacturing cyberinfrastructure relies heavily on human intervention to analyze complex data and execute experiments, which limits the pace and scalability of scientific discovery. This project addresses this issue by establishing a comprehensive training program to upskill the future workforce for Agentic Cyberinfrastructure Enabled Manufacturing Ecosystems. By transitioning systems from passive monitoring to active, autonomous decision making, Agentic AI has the potential to revolutionize how manufacturing research is conducted. The project will train 54 graduate and undergraduate students from diverse backgrounds nationwide, empowering them to become active creators of intelligent systems rather than passive users. This initiative serves the national interest by accelerating fundamental scientific innovation, enhancing national manufacturing competitiveness, and developing a highly skilled STEM workforce prepared to lead the future of intelligent factories. The technical goal of this CyberTraining program, led collaboratively by the University at Buffalo and the University of Georgia, is to equip researchers with the expertise to design, deploy, and govern AI agents within cyber physical manufacturing environments. The project scope encompasses an eight module curriculum that mirrors the full agentic manufacturing lifecycle. The methodological approach combines theoretical instruction with hands on learning using a cyber physical testbed. Trainees will integrate Large Language Models, Vision Language Models, and edge robotics to master the foundational Perceive-Reason-Act loop. The curriculum advances through generative design using Retrieval Augmented Generation, physical verification via Physics Informed Machine Learning to prevent AI hallucinations, multimodal inspection, and multi agent swarm orchestration. By integrating these advanced cyberinfrastructure technologies, the project contribution is the creation of a robust, scalable educational framework that transitions participants into capable architects of trustworthy, autonomous systems that can safely navigate physical constraints and accelerate advanced manufacturing research. 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.