OPAL: Orchestrated Platform for Autonomous Laboratories to Accelerate AI-Driven BioDesign (DRAFT)

The scaling of biodesign applications for the bioeconomy requires addressing critical challenges in modeling dynamic genomic, molecular, and environmental parameters that govern biological functions while overcoming limitations in current experimental and data collection methodologies. This proposal outlines an integrated strategy to establish an autonomous biodesign infrastructure under the DOE Orchestrated Platform for Autonomous Laboratories (OPAL) framework, leveraging robotic experimentation, AI agents, and standardized data architecture. Through three coordinated activities—autonomous protein engineering, microbial genome-to-function mapping and engineering, and adaptive phenotyping of plant systems, the project aims to create predictive, closed-loop workflows for functional discovery at multiple biological scales. As a scientific crosscut, research across the three activities focuses on understanding, design, and optimization of bioextraction/concentration of rare earth elements for critical US mineral supply chain advances. For protein engineering, Argonne National Laboratory (ANL) proposes a federated agentic robotics platform for engineering novel/optimized protein function for bioleaching of minerals, integrating dexterous humanoid robots with AI-guided design workflows. Lawrence Berkeley National Laboratory (LBNL) and Pacific Northwest National Laboratory (PNNL) will develop a distributed platform to map microbial genomes to functions using Pseudomonas putida produced organic acid complexants for rare earth element bioextraction. Oak Ridge National Laboratory (ORNL) will transform its Advanced Plant Phenotyping Laboratory (APPL) into an adaptive testbed equipped with AI agents to study Poplar species for enhancing mineral uptake. These efforts will be unified by interoperable data systems, standardized ontologies, and agentic AI frameworks, collectively advancing the DOE’s mission in predictive biodesign. By integrating high-throughput experimentation, FAIR (Findable, Accessible, Interoperable, and Reusable )–compliant data infrastructures, and cross-laboratory workflows driven by AI and robotics, the initiative seeks to accelerate autonomous discovery and provide transformative insights for the bioeconomy and critical mineral supply chain efforts.

Science Challenge

A key challenge in scaling biodesign applications for the bioeconomy lies in fully understanding and accurately modeling the dynamic genomic, environmental, and molecular parameters that govern desired biological functions, while simultaneously addressing limitations in current experimental and data collection methods. Manual, bench-scale experimentation and fragmented data systems currently cannot (1) generate the comprehensive, standardized datasets needed to capture the complexity of these parameters, (b) properly capture the dynamic and emergent functional outputs, and (c)  support advanced AI-driven modeling efforts required to address such complexity. To overcome these barriers, biology must integrate autonomous systems combining robotic experimentation, standardized AI-ready data/metadata ontologies and infrastructure, and intelligent agents capable of generating research hypotheses and orchestrating the execution of scientific workflows. Such systems would not only enable scalable and autonomous discovery but also provide the precision and throughput required to uncover actionable insights into the intricate interplay of biological variables that underpin predictive and scalable biodesign. These advancements will push the boundaries of DOE’s Biological and Environmental Research (BER) and Advanced Scientific Computing Research (ASCR) priorities in biological and autonomous sciences, paving the way for accelerated autonomous discovery and transformative contributions to the bioeconomy.