7 Teams Funded in Final Year of Scialog: Automating Chemical Laboratories

Research Corporation for Science Advancement, the Arnold and Mabel Beckman Foundation, and the Frederick Gardner Cottrell Foundation have made awards to 7 multidisciplinary teams of early-career scientists doing fundamental science to accelerate advances in automated laboratory technologies in the third and final year of the Scialog: Automating Chemical Laboratories initiative.

The 15 individual awards of $60,000 in direct costs will go to 15 researchers from a variety of institutions in the United States.   

During the three years of the initiative, RCSA and its funding partners made 50 individual awards totaling $3.3M. More than 90 individual Fellows participated in meetings throughout the three-year course of the initiative, creating a network that is poised to do decades of the breakthrough science that happens at the intersection of disciplines and perspectives.  

Scialog is short for “science + dialog.” Created in 2010 by RCSA, the Scialog format supports research by stimulating intensive interdisciplinary conversation and community building around a scientific theme of global importance.  

The final meeting held April 16-19, 2026, in Tucson, Arizona, brought together 45 early career scientists from disciplines including all areas of synthetic chemistry (organic, inorganic, materials and biological), integrated and automated instrument development, engineering, materials science, computer and data science, and AI computer research. Teams of two or three Fellows who had not previously collaborated wrote and pitched 22 proposals for seed funding for innovative projects they developed at the conference.  

In his welcoming remarks, RCSA President & CEO Eric Isaacs hearkened back to Thomas Edison as the archetype of the modern R&D scientist, one driven by persistence and trial and error more than instant insight.  

Isaacs noted that although Edison’s own path to the light bulb (thousands of filament tests, ending with carbonized bamboo) is often portrayed as the work of a lone genius, his success depended on large teams.  

“He believed in trial and error,” Isaacs said. “That was his method, and it required hundreds of people.”  

Though modern science has added theory and sophisticated conceptual frameworks to help reach answers faster than pure trial and error, Isaacs said much of today’s lab work is more Edisonian than we would like to admit. 

“AI and machine learning now offer the possibility not just of accelerated discovery, but of discovering molecules, materials, and behaviors we never would have imagined,” he said. “But the bottleneck to progress may be in integrating these new tools given the constraints of experimental work.”  

“AI is a wonderful tool, but in the end we’re humans,” he said. “You’ve got to make stuff.”  

Keynote speaker Christopher J. Welch, Indiana Consortium for Science and Engineering, connected prediction and experiment as the core loop of autonomous systems in his talk, “Advancing the Automation of Chemical Laboratories through Industry-Academia Research Collaborations.”  

In presenting case studies from projects carried out over two decades at Merck Research Laboratories and lessons learned from recent projects at the NSF Center for Bioanalytic Metrology, an industry-university cooperative research center, he described much of his career as building “problem-solving engines.”  

“You can string them together and create really powerful tools,” he said, predicting a future where autonomous, AI‑guided systems that generate predictions, run experiments via automated platforms, and iteratively refine solutions.  

He said the human role will shift more toward designing the engines and interpreting higher‑level strategy, not running individual experiments.  

Welch also argued that the future belongs to teams and networks, not lone labs.  

“I know what you’re thinking: I can do 3D printing, I’m good with a soldering iron, I can do Python coding,” he said. “But do-it-yourself is a real trap. Before you know it, you’re doing some stuff that is really not that relevant to your company or your department’s objectives.”  

He urged academic researchers to use external platforms instead of re‑building infrastructure, co‑develop tools with academia and industry partners, and to treat collaboration as both a strategic advantage and a source of personal satisfaction.  

“I’ve had a lot of fun through collaboration,” he said. “It’s allowed me to expand what I can do.”  

In addition to Welch, other senior scientists guiding discussions at the meeting as Facilitators were: Rajeev Surendran Assary, Argonne National Laboratory; Lane Baker, Texas A&M University; Malika Jeffries-EL, Boston University; Anne LaPointe, Cornell University; Philip LeDuc, Carnegie Mellon University; Karl Mueller, Ames National Laboratory; Nikki Pohl, Indiana University, Bloomington; and Sarah Reisman, California Institute of Technology. 

The meeting included brief presentations by last year’s awardees, who reported on their teams’ progress and challenges. For example, one team is working on theoretical and experimental methods for automated phase-space exploration in materials discovery, repurposing ideas from industrial statistical process control (the kind of rules used to decide when to shut down a factory line) to monitor automated chemical experiments in real time. Another group is working on temporally adaptive, AI‑driven optimization of organic redox‑flow batteries.  

The conference also included a session on the Genesis mission, a new initiative by the Department of Energy to build the world’s most powerful AI-driven scientific platform. Scialog topics (AI-driven autonomous labs, data‑intensive experiments, quantum, energy materials, critical materials) sit directly inside Genesis’ three pillars: energy innovation, discovery science, and national security. An initial $290 million funding call was issued in mid-March, and future calls will focus on AI advantage and team collaboration, involving both national labs and universities.  

Isaacs, who serves on the DOE Office of Science Advisory Committee (SCAC), the conduit between the research community (academia, labs, industry) and the Office of Science helping shape science strategy and priorities.  

He advised Fellows looking to apply for future Genesis funding to make sure their proposal isn’t just adding AI to an existing project.  

“Anything you propose just has to clearly articulate what you can’t do now that you will be able to do because you have whatever tool you choose to use in AI.”  

Mueller, Director of Ames National Laboratory, is one of the 17 DOE national lab directors and has served on the National Lab Directors Council, which co‑conceived the Genesis mission with DOE.  

He said Fellows at the conference were ideating in the same problem space Genesis is paying for, so the ideas and collaborations born at Scialog would be well aligned to be turned directly into competitive Genesis proposals.  

He urged potential Genesis applicants to think through how they define and measure AI advantage in proposals.    “Am I just using AI because it’s cool, or am I using AI because I can actually get to something that will have a bigger impact in a faster period of time?”  

Assary, Group Leader/Chemist Molecular Materials Group at Argonne National Laboratory, said Genesis is built around platforms at DOE labs and external partners, not solo PI projects. He said new data and models from these projects will become part of DOE’s shared ecosystem and advised designing projects to use platforms such as ModCon and the American Science Cloud rather than relying solely on local resources. 

RCSA Senior Program Director Andrew Feig, who led the initiative, emphasized that innovation at scale requires aligning synthesis, analysis, and computation so scientists can capitalize on predictions instead of stockpiling them.  

“If your calculations are putting out 10,000 predictions a day or a million predictions a day, yet your synthesis can do 100, you’ll have a mismatch,” he said. “When your whole pipeline is balanced enough to turn those predictions into real discoveries, you’ll really hit paydirt.”   

Scialog’s intensely collaborative format is exactly the kind of environment that can stitch those pieces together across different teams so the whole community moves faster than any single lab could alone, he said.  

The following Scialog: Automating Chemical Laboratories teams will receive 2026 Scialog Collaborative Innovation Awards: