PCL-test bed: the artificial intelligence for materials design network (AIMD-net)

The National Science Foundation awarded The Johns Hopkins University a $5 million cooperative agreement on August 1, 2026, under the NSF Technology, Innovation, and Partnerships program to develop the AI for Materials Design Network (AIMD-Net), a national test bed of programmable cloud laboratories combining automated robotic experimentation with distributed artificial intelligence.

AIMD-Net addresses the challenge of scaling autonomy from individual instruments and laboratories to achieve coordinated AI-driven decisions and real-time information flow across a network of cloud laboratories. The project delivers a data infrastructure and information exchange layer comprising three integrated components: a scalable, event-driven cyberinfrastructure for real-time cross-site data and information exchange; a modular multi-agent decision fabric using shared uncertainty quantification schemas to support information flows between AI decision makers and agentic workflows; and a semantically complete exchange protocol for data standardization. The network will accelerate development of structural materials for extreme environments in defense, transportation, and energy applications and train the next generation of scientists and engineers in AI-driven research methodologies.

Performance occurs at Baltimore, Maryland from August 1, 2026 through July 31, 2030. The funding agency obligated $5 million to the award.

Abstract

PCL-test bed: the artificial intelligence for materials design network (AIMD-net) flexible networks of autonomous ?self-driving? Laboratories, which combine automated robotic experimentation with distributed artificial intelligence (AI) will revolutionize science and engineering by vastly increasing the rate at which knowledge is generated. The AI for materials design network (AIMD-net) will address the challenges of scaling autonomy from individual instruments and labs to achieve real-time information flow and coordinated AI-driven decisions spanning a national test bed network of programmable cloud laboratories. AIMD-net?S robust, secure, and novel data infrastructure will create AI-ready data with unprecedented speed to enable training advanced AI models and accelerate development of structural materials for a broad range of conditions including extreme environments encountered in defense, transportation, and energy applications. In SUM, AIMD-net will help secure the national defense, enhance economic competitiveness, and train the next generation of scientists and engineers with the skills necessary to advance the frontier of AI-driven research. Building an AI-driven autonomous network of programmable cloud laboratories requires solving critical challenges at all levels of experimental investigation, from the instrument level to coordination across the lab to full integration into a network of labs. AIMD-net will address challenges across all these levels with a novel data and information exchange layer of three integrated components: (I) a scalable, event-driven and interoperable cyberinfrastructure for real-time cross-site data and information exchange? (II) a modular multi-agent decision fabric using a shared uncertainty quantification schema supporting information flows between AI decision makers and agentic workflows? And (III) a semantically complete yet lightweight exchange model for data, workflows, and authorization within and between network nodes. To facilitate AI-driven autonomous operations, multiple AI agents will interact across this data infrastructure and cooperate at scales ranging from individual instruments to the entire network. AIMD-net will establish a hierarchy of agents using a standardized computational and communication infrastructure to enable AI-driven uncertainty-aware and optimal design workflows automating experimental execution, coordination across instruments, aligning with human-prompted scientific objectives to ensure secure, robust AI and human interoperability across the system. AIMD-net will transform materials research through two science campaigns: data-driven multi-objective materials optimization of alloys for extreme conditions, e.g., high temperature, shock loading, and aggressive oxidizing/plasma atmosphere, and development of the first data-driven scientific foundation model (SCIFM) for materials. AIMD-net will also provide a powerful training and workforce development pipeline for autonomous agentic workflow development. The project will build a community of innovation across academia and industry that drives autonomous materials design and discovery and advances AI-driven distributed autonomy more generally. 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.- subawards are planned for this award.