The Department of Energy's Oak Ridge National Laboratory has launched a novel robotic platform to rapidly analyze plant root systems as they grow, yielding AI-ready data to accelerate the development of stress-tolerant crops for new fuels, chemicals and materials.
The new platform adds belowground imaging to ORNL's Advanced Plant Phenotyping Laboratory (APPL) , an automated facility that already uses high-resolution cameras to quickly assess aboveground plant traits. The combination of information gathered from the root system combined with data collected on plant traits aboveground enables faster and deeper insights into the connections between the form and function of roots in the soil and plant productivity and hardiness above. With this one-of-a-kind capability, APPL can now repeatedly image plant roots growing in soil without disturbing the plant.
Massive datasets generated in APPL are analyzed using AI and ORNL's Frontier exascale supercomputer, part of the Oak Ridge Leadership Computing Facility, a DOE Office of Science user facility. APPL's capabilities support the goals of the DOE Genesis Mission to accelerate science through AI innovation.
The hidden foundation of plant success
Root systems are the hidden foundation of plant success, regulating water and nutrient uptake, stress response and vital interactions with the soil environment.
"If you're only looking at the aboveground plant, you're only getting half the story," said Larry York, ORNL senior scientist who specializes in studying roots and works closely with APPL.
Roots are difficult to observe directly, historically requiring invasive, labor-intensive sampling methods that limit real-time insights.
To generate root data more quickly, APPL scientists developed a robotic imaging platform using rhizoboxes - clear, flat boxes containing plants growing in soil and equipped with a sliding cover to block light. As each rhizobox moves into an APPL imaging chamber, the cover is robotically removed and the root environment is imaged using a high-resolution color camera and a near-infrared camera. The system operates autonomously 24/7 over the course of an experiment.
The system generates large amounts of data as plants and their roots grow, including information on traits such as root length and diameter, nutrient and mineral uptake, and water content of the roots and surrounding soil.
"With APPL's enhanced capabilities, we can for the first time simultaneously link aboveground and belowground performance in the same plant at the same time across multiple modalities," said Jerry Tuskan, director of the DOE Center for Bioenergy Innovation and Corporate Fellow at ORNL. "We can look at, for instance, photosynthesis function aboveground and water uptake belowground, enabling associations between plant anatomy and function above and below. That is a unique capacity of APPL; no other facility in the world can do that."
APPL root imaging and analysis at a glance:
- 500-plus rhizobox capacity
- 30x40-centimeter viewable window
- Variable box angle to encourage visible root growth
- Variable watering depths - top, middle and bottom
- 24/7 robotically controlled imaging
- High-resolution color camera for structural root measurements
- Near-infrared high-resolution camera for imaging physiological data such as water uptake
- Up to 1 terabyte of data collected per week on roots and shoots
- AI-assisted data analytics
APPL's belowground imaging system "gives us a perspective that we can't get easily today in the field, where we're limited to a sample collection at a defined point in time," Tuskan added. "In APPL we get a time series perspective and an autonomous sampling method that leaves the plant and its roots intact for future study."
Certain features such as the ability to water at different depths can simulate various growing scenarios as well, Tuskan said. Watering from below simulates conditions in which groundwater elevates the water table, a time when plants get their water from below. Watering from above mimics how water percolates down through the soil.
The use of automation, advanced imaging and AI enable scientists to rapidly answer such questions as whether plants are investing more in roots or shoots, a basic question in plant science, York said. APPL's analytics and the combination of below- and aboveground data will let scientists examine such ratios across different plant genotypes and soil treatments to answer questions such as how plants are responding to drought, and which roots support colonization of plants by beneficial soil microbes. Integrating whole-plant data allows new perspectives on how shoots and roots work together to drive plant performance.
Time series, whole-plant analysis enable deeper insights
"You're not just doing a one‐off destructive harvest," York said. "By imaging roots in situ [in place], the system captures the behavior of the entire plant over time, enabling assessment of traits such as lateral root formation or overall root shifts as water availability changes. You're getting a profile of root system architecture, root density and how that relates to processes like water uptake."
For researchers such as Melanie Mayes, an ORNL Distinguished Scientist who focuses on how belowground resources such as root distribution, soil moisture, nutrients and microbes influence aboveground processes, APPL's root capability is a welcome advancement.
"Plant growth is informed by what roots are doing and our current capacity to see roots is extremely limited, representing only a snapshot in time," Mayes said. "What APPL enables is deep insight into how plants are taking up things like water, nutrients and can even act as hyperaccumulators of critical minerals." APPL could also help scientists better understand soil pore structure - which determines how well soil can store and transmit water, air and nutrients - by running experiments using soil collected from field sites of interest.
APPL advances development of AI-guided, self-learning lab network
APPL's capabilities are being leveraged in a multi-institutional DOE project called OPAL, or Orchestrated Platform for Autonomous Laboratories . OPAL scientists are creating an interconnected network of self-learning labs at four national laboratories using AI, robotics, automated facilities and a common data platform to accelerate biotechnology breakthroughs. The OPAL project is funded by the DOE Office of Science Biological and Environmental Research program and the Office of Advanced Scientific Computing Research.
APPL and OPAL are aligned with DOE's Genesis Mission to develop an integrated discovery platform connecting the nation's supercomputers, experimental facilities, AI systems and unique datasets. Genesis targets a doubling of the productivity and impact of American research and innovation within a decade.
"APPL fundamentally shifts how researchers everywhere can study belowground and aboveground biology and understand holistically how plants are balancing their investments between roots and shoots," York said. "What we've built can improve the research pipeline for plant science by providing access to the facility and to open-source software. That's the real value of APPL."
APPL is available for scientific community use via research collaboration. For more information, see the APPL website , where interested users can submit a brief inquiry .
UT-Battelle manages ORNL for DOE's Office of Science, the single largest supporter of basic research in the physical sciences in the United States. The Office of Science is working to address some of the most pressing challenges of our time. For more information, visit energy.gov/science . - Stephanie Seay
