Three major research institutions are deploying artificial intelligence and robotics to transform how scientists discover and design new materials, potentially cutting years off the traditional research timeline. The City University of New York (CUNY) received an $18.1 million National Science Foundation (NSF) award to build an AI-powered materials lab, while the University of Pittsburgh is launching a Department of Energy-backed project to automate quantum computing research. These initiatives represent a fundamental shift in how materials science research happens, moving from human-led experiments to AI agents that design, run, and learn from thousands of experiments simultaneously.

What Is AI-BIOME and How Will It Change Materials Discovery?

CUNY's new platform, called AI-BIOME (AI-Enabled Bio-Inspired Materials Ecosystem), combines laboratory robotics, advanced instruments, shared data systems, and AI agents into a single research platform. The four-year project will give researchers nationwide remote access to automated tools for developing bio-inspired materials, from sustainable plastics to drug-delivery systems and optical coatings.

Unlike traditional high-throughput screening, which tests one combination at a time, AI-BIOME uses machine-learning-guided rapid exploration to test vast combinations of ingredients and processing conditions simultaneously. Researchers define the problems and evaluate findings, while AI agents analyze results, recommend promising next steps, and automatically carry out approved experiments. This approach is particularly powerful in materials science, where the sheer number of possible molecular interactions makes exhaustive human experimentation impossible.

"AI-BIOME will create a new kind of scientific discovery laboratory in which AI agents directly coordinate robotic synthesis, processing, and characterization equipment and decide which experiments to run next," said Ronald Koder, professor of physics at City College of New York and the project's principal investigator.

Ronald Koder, Professor of Physics at City College of New York

The NSF funding will enable major upgrades to CUNY research facilities. At City College of New York's Center for Discovery and Innovation, a high-throughput chemistry facility will be equipped with robots that handle liquids and solids to create and study polymers, hydrogels, colloids, biomolecular structures, and composite materials. At the CUNY Advanced Science Research Center, automated surface science, imaging, and photonics facilities will measure materials' structure, chemistry, strength, and optical, electrical, and thermal properties.

How Will AI Agents Automate Quantum Computing Research?

At the University of Pittsburgh, a team led by School of Computing and Information Professor Youtao Zhang is tackling a different frontier: measurement-based quantum computing. This approach uses dense, interconnected networks of qubits rather than isolated pairs, offering several advantages but presenting enormous computational challenges. The process of measuring these qubits requires imposing amounts of computing power, and not enough algorithms and principles have been developed to make use of any data collected.

Zhang's team is designing specialized AI agents to perform different research tasks automatically. The system includes agents to determine which measurements to make, develop new principles that can be tested, optimize the process for hardware, and manage the other three agents. By automating the workflow, the project aims to accelerate discoveries in chemistry, materials science, and energy.

"We're building AI that can automatically design, optimize and verify measurement-based quantum computing, automatically discovering better ways to build quantum computations," said Youtao Zhang, professor at the University of Pittsburgh's School of Computing and Information.

Youtao Zhang, Professor at University of Pittsburgh School of Computing and Information

The Pitt project includes collaborators at Virginia Tech and Argonne National Laboratory. The researchers are betting that AI agents are well-positioned to boil down the complexities of quantum research in ways that would be impossible for human teams. The system is designed to improve with time and experience, with the goal of creating and sharing a pipeline for measurement-based quantum computing that can contribute to future breakthroughs.

How to Access and Participate in AI-Powered Materials Research

• Remote Access: Researchers from universities, national laboratories, small businesses, and industry will be able to design and run experiments remotely through AI-BIOME, extending access to automated, AI-enabled laboratories beyond large corporations and well-funded institutions.
• Pilot Projects: Seven pilot projects will test real-world applications spanning sustainable materials, medicine, manufacturing, and AI-enabled characterization, including automated data generation for peptide assembly, water-responsive material manufacturing, and drug nanoparticle development.
• Workforce Training: AI-BIOME will create a teaching laboratory where students gain hands-on experience with laboratory automation and AI-assisted research, including modular certifications, internships, and outreach for high school, trade school, undergraduate, and graduate students.
• Shared Data Systems: A unified data system will connect each stage of materials research, from molecular makeup and formulation to processing and performance, capturing results in real time so AI systems can adjust experiments automatically.

The broader research infrastructure supporting these projects extends beyond individual institutions. The Massachusetts Green High Performance Computing Center (MGHPCC) in Holyoke, Massachusetts, provides state-of-the-art computing resources that enable millions of virtual experiments each month across climate, energy, medicine, materials science, health, and computing. MGHPCC-supported researchers are already using artificial intelligence and materials modeling to design catalysts and new materials for sustainable energy production, storage, and conversion.

Why Does This Matter for Scientific Discovery?

These AI-powered platforms address a fundamental bottleneck in materials science: the discovery space is too large for humans to explore experimentally using traditional approaches. By automating the design, execution, and analysis of experiments, AI agents can explore combinations that human researchers might never consider, potentially uncovering materials with properties that traditional approaches would never discover.

The economic implications are significant. Historically, research activities like those supported by MGHPCC have enriched the economy by yielding new companies and sometimes entirely new industries. By accelerating the time required to move promising technologies from the laboratory into practical use, these AI-powered platforms could compress years of research into months.

"This is an incredibly exciting moment for CUNY and for materials science. Artificial intelligence is transforming how we discover, but it needs laboratories that can learn and experiment alongside us. With AI-BIOME, we're creating exactly that capability, opening new opportunities for researchers, students, startups, and industry to innovate together," said Rein Ulijn, director of the CUNY Advanced Science Research Center Nanoscience Initiative and distinguished professor of chemistry at Hunter College.

Rein Ulijn, Director of CUNY Advanced Science Research Center Nanoscience Initiative

The NSF's Test Bed: Toward a Network of Programmable Cloud Laboratories program selected AI-BIOME as one of 20 inaugural projects, signaling federal recognition of this research model's potential. Similarly, the Department of Energy's Genesis Mission, which supports Pitt's quantum computing project, represents a historic national initiative to build the world's most powerful integrated science discovery platform by uniting government, industry, academia, and philanthropy.

These investments reflect a broader shift in how scientific research is conducted. Rather than individual researchers working in isolation, AI-powered platforms enable collaborative, distributed discovery where human expertise and machine learning work in tandem. The result is faster innovation, broader participation, and the potential to solve scientific problems that have remained intractable for decades.

" }