CAREER: BRACE-CI: Building Resilience through human-AI Co-Engineering Cyberinfrastructure

This Faculty Early Career Development project, BRACE-CI, creates new ways for engineers and artificial intelligence (AI) to work together to transform the design and assessment of the next generation of civil infrastructure. Communities face escalating challenges from natural hazards, aging infrastructure, increasing system complexity, and environmental impacts, yet fragmented engineering workflows and disconnected information across researchers and stakeholders limit the ability to evaluate and implement safer, more efficient, and more resilient solutions. BRACE-CI addresses this challenge by developing an open cyberinfrastructure that brings human judgment and AI into a shared engineering environment. The project helps engineers assess structures more rapidly and comprehensively, explore integrated design options from concept through detailed design, and make better decisions about safety, resilience, cost, and resource efficiency. BRACE-CI prepares students and professionals to use human-aligned engineering AI responsibly in safety-critical settings. Through open educational and research resources and outreach activities, the project broadens access to advanced engineering tools and contributes to building resilient communities and the STEM workforce. BRACE-CI advances human-aligned co-engineering through master builder agents: AI systems that assist engineers across inspection, design, analysis, and coordination tasks while preserving human oversight and improving civil infrastructure outcomes. The research is organized around three connected thrusts. Co-INSPECT creates realistic virtual worlds for civil infrastructure and uses AI assistance to support rapid, explainable, and more complete assessments. Co-DESIGN creates a multidisciplinary design environment that helps architects, engineers, and builders examine coordinated building options across conceptual, preliminary, and detailed design stages. Co-TEACH brings the research platform into classrooms, mentored research experiences, and workshops, while using feedback from students and practitioners to improve the platform and guide future research. Across these thrusts, BRACE-CI integrates engineering simulations, domain knowledge, multimodal AI agents, and scalable computing resources. The resulting reproducible cyberinfrastructure supports both recursive self-improvement of agents and operation by humans in the loop. The resulting methods, educational resources, and research platforms advance interpretability and safety in AI for safety-critical engineering, accelerate resilient infrastructure design and assessment, and prepare the next generation of scientists and engineers to lead responsible human-AI collaboration. 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.