Collaborative Research: Human-centered AI for Reconfigurable Manufacturing Operation and desigN sYnthesis (HARMONY)

About This Grant

This award supports research that focuses on the development of Human-centric AI for Reconfigurable Manufacturing Operation and desigN sYnthesis (HARMONY) to harmonize human workers with increasingly autonomous resources. Emerging technologies, such as automated guided vehicles, mobile manipulators, collaborative and reconfigurable robots, offer new opportunities to address persistent challenges like machine breakdowns, material shortages, and demand fluctuations through dynamic reconfiguration of hardware, software, and logistics. However, fully realizing these benefits requires advanced analytical and programming skills to orchestrate multiple autonomous resources while meeting multi-facet performance targets and operational constraints. While generative artificial intelligence (GenAI) offers an intuitive interface for human-autonomy collaboration, its current application in manufacturing is limited by a lack of domain-specific knowledge. This project seeks to create preference-aligned decision options that humans can explore, select, and refine through low-barrier, multimodal interfaces. The project looks to also include an educational program featuring an innovative curriculum in AI and digital manufacturing, hands-on K–12 engagement and professional development opportunities delivered via workshops, webinars, industry partnerships, and Smart Learning Factories. Successful implementation of HARMONY has the potential to transform manufacturing systems to incorporate autonomous resources in rebuilding national manufacturing capacity and prepare future workforce with advanced technologies. The goal of this research is to innovate GenAI solutions along with digital twins, and progressive data cultivation strategies for translating the autonomy of individual resources into system-level performance gain, all while under human supervision. Existing GenAI models face significant limitations due to a lack of domain specificity, limited manufacturing decision data, and the risk of hallucination. To overcome these challenges, this project pursues three key objectives: (i) the development of a progressive data cultivation strategy leveraging digital twins and statistical surrogate models; (ii) the design of tailored GenAI architectures that embed manufacturing-specific constraints, objectives, and interdependencies; and (iii) the systematic evaluation through Smart Learning Factories and industry collaborations. This project aims to advance the next generation of GenAI technologies for human-centric decision-making in dynamic, reconfigurable manufacturing environments. 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.