Biologists and engineers have been working together over decades to decipher the language of biology. Now, they are teaching AI models to interpret that knowledge and harness it for national priorities through the Orchestrated Platform for Autonomous Laboratories (OPAL) project, a Department of Energy (DOE) investment tied to the Genesis Mission.

OPAL researchers are building an entirely new research ecosystem for biological research to meet DOE missions. Hands-on, manual experimental cycles are being replaced with intelligent AI-driven systems that integrate historical data with fresh research results to learn, adapt, and optimize in real-time. 

An orchestra for biotechnology

Just as a musical orchestra requires four instrument groups—strings, woodwinds, brass, and percussion—to perform a symphony, the orchestrally named OPAL requires contributions from four distinct sources to create a new kind of biotechnology powerhouse. 

Pacific Northwest National Laboratory (PNNL) plays its essential part in microbial research and autonomous experimentation. PNNL co-leads OPAL’s microbial testbed in partnership with Lawrence Berkeley National Laboratory (LBNL), focusing on microbial bio-design and phenotyping—the assessment of an organism’s observable traits. While PNNL undertakes a broad range of microbial research and autonomous experimentation, LBNL is focused on microbial bio-design and assessing observable microbial traits. 

At PNNL, the microbial testbed team takes advantage of the new Anaerobic Microbial Phenotyping Platform (AMP2), a powerful new resource unique among the national laboratories. AMP2 became operational in January 2026 within the Environmental Molecular Sciences Laboratory, a DOE Office of Science user facility at PNNL. This sophisticated platform integrates robotic systems that automate formerly manual steps of preparing microbial samples, transferring experiments between stations, and timing sample incubation. 

But robotic automation is just the start. True autonomy comes from the integration of control software, automated data analysis, and intelligent AI agents executing continuous cycles of experiment without human intervention. Within OPAL’s Design-Build-Test-Learn loop, AMP2 performs the heavy lifting. Meanwhile, advanced instrumentation located at PNNL enables the large-scale identification of proteins, bridging the gap between genes and the proteins that lead to biological function. 

“PNNL’s world-class measurement capabilities in mass spectrometry, proteomics, and metabolomics form one of the cornerstones of OPAL’s biological design platform,” said Kristin Burnum-Johnson, PNNL’s OPAL project lead. “Our contribution includes generating those data and making them accessible to project-developed AI agents for rapid interpretation and systems integration.”

OPAL FAMOUS

When learning a new language, it helps to have an interpreter. For an AI agent to understand the language of biology, that means training it with science research outputs and advances published in the scientific literature. It also means creating models that “translate” science inquiries across the “dialects” of instrument data outputs, gene and protein sequences, and organism growth patterns, among other data points. 

A research team at PNNL is doing just that through a separate but connected project supported by DOE’s Advanced Scientific Computing Research (ASCR) program called OPAL Foundational AI Models for Optimizing and Understanding Biological Systems (FAMOUS).

“For biology, one of the most underappreciated and difficult tasks in AI training is finding data that’s in the right format,” said Chris Oehmen. “Through OPAL FAMOUS, we are creating agents that ingest raw data and translate it into terms that an AI agent can understand and then act upon.”

When implemented as part of the OPAL platform, these agents will act as the “central nervous system” of advanced automated laboratory platforms.

PNNL’s expertise in microbiology, predictive phenomics, and purpose-built AI agents for biology provide one of the four orchestral contributors to OPAL. 

Oak Ridge National Laboratory brings its Advanced Plant Phenotyping Laboratory, which is unique among the national laboratories, to the overall OPAL effort through a synergistic partnership with PNNL’s expertise in microbial systems. Argonne National Laboratory contributes extensive experience in protein structure and data analyses to inform protein design. And LBNL’s genomics and bio-design platform connect microbial functions with genomes in partnership with PNNL. 

Together, the four labs are training their combined expertise to deliver on two DOE Genesis Mission Challenges: improving bioproduct production and critical mineral extraction using microbes. 

Nature’s factories

Plants and microbes already provide efficient bio-factories, churning out products that we use every day. Cotton, olive oil, antibiotics, dyes, lubricants, fuels, and thousands of other commodities trace their origins to plants or microorganisms.

These efficient bio-powerhouses are now being harnessed to churn out other commodity chemicals, in collaboration with industry partners through the OPAL project. 

“Harnessing these processes would provide an alternative to a supply chain that is dependent on China, all while advancing U.S. leadership in the bioeconomy,” said Douglas Mans, interim Associate Laboratory Director of Earth and Biological Sciences at PNNL. “What’s been holding us back is our inability to grow and study the massive number and diversity of microbes at a scale and pace that is needed for identifying breakthrough applications. With the work we are doing in OPAL, we will be able to identify, grow, and optimize the use of these microbes in days and weeks instead of years using automation and AI, a focus of the Genesis Mission.” 

OPAL researchers are leveraging AI insights to learn the underlying forces and biomolecules that control biological function and in turn how to use this information to predict and design organism growth for higher yields or better stress tolerance.

Recovering critical minerals

The team is currently tackling a DOE priority challenge to optimize growth of Thlaspi arvense, commonly called field pennycress. The plant has already shown promise to serve as a field cover crop that can produce useful oils while pulling the critical mineral nickel out of the soil.

Likewise, soil microbes such as Pseudomonas putida have shown promise in “biomining”—the practice of using organisms’ natural ability to break down minerals bound up in rocks in a process that releases the desirable rare earth metals. In its first two months of operation, AMP2 has already identified key biological pathways to optimize critical mineral recovery under preferred laboratory growth conditions. 

“We are excited to move this work toward optimizing both microbes and plants, because microbes and plants work well together—you can use either one to make the other one better,” said Oehmen. 

“OPAL is teaching AI the language of biology, then using it to steer growth and production,” added Burnum-Johnson. 

“This is about much more than simply making current processes faster and more efficient,” said Mans. “Automation and AI are vehicles for true scientific innovation. We can perform many more experiments and generate much larger datasets that will lead to new insights that we cannot even imagine.”