Between 5 and 25 per cent of the harvest is lost when stalks fall - solving stalk strength with engineering, genomics and AI (NSF award $2M)
The U.S. National Science Foundation awarded $2 million to collaborative research including Clemson University to understand why the stalks of corn, wheat and sorghum break and topple before harvest, combining engineering, plant biology, genetics and AI.
Grant overview (primary data)
- Award amount$2,000,000
- RecipientClemson University (South Carolina)
- ProgramEPSCoR RII: Focused EPSCoR Col
- Period2026-08-01 〜 2030-07-31
- FunderU.S. National Science Foundation (NSF) / NSF
Key points
- Lodging destroys 5 to 25 per cent of the potential harvest and costs farmers billions of dollars each year.
- The genome-to-phenome bottleneck is that millions of genetic markers vastly outnumber the plants that can be measured.
- The aim is shifting from correlative association toward predictive, mechanism-based design grounded in engineering physics and biology.
- It combines engineering, plant biology, genetics and AI, automating high-throughput phenotyping and building genomic foundation models.
- Harvest-integrated sensing is advanced toward field deployment, carrying results into a breeder-ready toolkit.
- Of the 120 NSF awards this site holds as of 2026-09-02, obligated amounts run $1M to $11M with a median of about $1.96M.
1The loss called lodging
Stalks breaking and toppling before harvest is called lodging. According to the abstract it destroys 5 to 25 per cent of the potential harvest, wasting food and costing farmers billions of dollars. Given that cereals supply most of the calories people eat worldwide, a loss occurring just short of harvest bears on food supply along a path separate from planted area or yield per acre.
2More to ask about than can be measured
The abstract calls this the genome-to-phenome bottleneck. Moving from breeding that relies on correlative association toward predictive, mechanism-based design grounded in engineering physics and biology is how the award frames its approach: not gathering more correlations but attacking from the side of why a stalk breaks.
3What three disciplines were seeing separately
- 1EngineeringTreats it as a problem of mechanics - why the stalk fails
- 2Plant biologyLooks at how cellular and subcellular structure sets strength
- 3GeneticsLooks at which genetic factors bear on strength
- 4AIBinds data of differing kinds into multi-modal genomic foundation models
The award sets out automating high-throughput phenotyping, resolving the cellular and subcellular drivers of stalk strength, and developing multi-modal genomic foundation models. It also advances harvest-integrated sensing toward field deployment, carrying results as far as a breeder-ready toolkit.
4Scale and reach of the award
Of the 120 NSF awards this site holds as of 2026-09-02, the obligated amounts run from $1 million to $11 million with a median of about $1.96 million. The $2 million here sits just above the median.
It was made under EPSCoR, a framework whose purposes include building research capacity outside regions where research funding concentrates; the award also speaks of training students at community colleges and less research-intensive institutions.
Why it matters
What is lost before harvest bears on supply along a path separate from planted area and yield. A tool that predicts stalk strength in a breeder's hands could reduce the number of breeding cycles required. Building sensing into harvest machinery turns the harvest itself into an occasion for data collection, which concerns machinery makers and seed companies alike.
FAQ
What is lodging?
What is the genome-to-phenome bottleneck?
What is AI used for?
Sources (primary)
Source: NSF Award Search (U.S. National Science Foundation, public domain). Amounts are the obligated amount. For privacy, we do not handle principal investigator names.
- NSF Award (original, official)
- NSF Award ID: 2614823