Globus Platform Will Enable Autonomous Laboratories for Materials Research
August 11, 2026 | Vas Vasiliadis
We’re excited to share that Globus will serve as enabling technology in two awards recently announced by the National Science Foundation, focused on creating remotely accessible, AI-driven laboratory facilities for materials research.
PoLARIS: A Cloud Laboratory for Soft Materials
The Polymer Laboratory for AI, Robotics, Informatics, and Standards (PoLARIS) received $20 million to build a remotely operated, robotic laboratory for polymers and other soft materials. The project is led by the University of Chicago Department of Computer Science and Pritzker School of Molecular Engineering, and Argonne National Laboratory, with co-principal investigators Ian Foster, Jie Xu,Stuart Rowan, and Yuxin Chen.
PoLARIS will link more than 20 autonomous stations across five facilities: UChicago’s Hyde Park High-throughput Polymer lab, Soft Matter Characterization Facility, and Materials Preparation and Measurement Laboratory, plus Argonne’s Advanced Photon Source and Polybot. Researchers anywhere in the country can design and submit experiments to the lab via cloud interfaces rather than traveling to the instruments. At scale, the team expects to run up to 300 experiments per day, with end-to-end autonomous experiments for the initial science drivers targeted within 18 months and external user service by the end of year two. PoLARIS is one of 20 nodes in NSF’s Test Bed: Toward a Network of Programmable Cloud Laboratories initiative, a $380 million investment announced in July 2026.
Our co-founder, Ian Foster, describes the result as “essentially a distributed operating system for scientific discovery”, with instruments in different facilities and institutions coordinated by AI agents so they behave as one laboratory. He notes that the infrastructure principles underpinning the Globus platform, such as reliable task orchestration and federated access control, apply equally to directing physical experiments remotely.
A New Materials Innovation Platform
NSF also committed $50 million to two new Materials Innovation Platforms, each receiving $25 million over six years: one at Texas A&M University, which will combine robotics and AI in an autonomous laboratory for alloy discovery, and another at the University of Wisconsin, which will study how chemical and structural complexity affects material performance in extreme environments.
Both platforms operate as national user facilities, dedicating a majority of instrument time to external users and supporting AI-driven autonomous experimentation. Globus will contribute to the University of Wisconsin project and provide core enabling infrastructure for data management and computation.
Shared Infrastructure Requirements
Different science, familiar plumbing. Facilities like PolARIS and the MIP need to move instrument data reliably to compute and archival systems; authorize external users without provisioning local accounts for each of them; orchestrate multi-step workflows that span instruments, robots, HPC, and AI models under varying administrative controls; and capture sufficient provenance to enable data publication and research reproducibility. Solving these challenges is at the core of what the Globus platform has been doing for almost two decades. Globus services such as fast file transfer, secure data sharing, federated authentication, remote computation and reliable orchestration are key reasons the platform ends up in the critical path of robotic laboratories and other automated research scenarios. We look forward to enabling the next frontier in automated discovery!