Creating robust crops for millennia depended on human or natural pollination, which made the process time-consuming and sometimes expensive.
The technology, called Genome Editing with Artificial Intelligence-based Robots (GEAIR), solves a long-standing pollination problem: hybrid breeding, which includes cross-pollination between distinct parent plants, results in higher-yielding, hardier crops.
However, it relies on slow, expensive manual labor to handle pollination since recessed stigmas (female organs) and intricate floral structure in major crops such as tomatoes and soybeans prevent robotic automation.
Human hands can navigate these flowers, but at a steep price.
Cao Xu, Study Corresponding Author and Professor, Institute of Genetics and Developmental Biology (IGDB), Chinese Academy of Sciences
Manual pollination in China accounts for more than 25% of fresh-market tomato breeding expenditures. Emasculation, or the removal of male parts to prevent self-pollination, accounts for 40% of that labor. Soybeans’ tightly sealed flowers prevent natural cross-pollination, necessitating such labor-intensive manual breeding that farmers are still unable to reap over 30% yield boost from hybrid vigor.
Drawing inspiration from the Green Revolution, which reengineered crops for equipment, the team pioneered “crop-robot co-design” to build crops that are well-suited to robotic technology.
On the crop design side, they employed CRISPR-Cas9, a precision gene-editing technology, to target B-class MADS-box genes like GLO2 in tomatoes, which control flower development. The end result produces plants that are both male-sterile (no need for emasculation) and have protruding stigmas, making them easier for robots to reach.
Xu added, “We gave flowers a makeover for machines.”
To complement the newly engineered crops, GEAIR's bespoke robot has deep learning-based computer vision that recognizes ready blooms and utilizes a precision arm to drop pollen, matching human efficiency while operating around the clock.
Its versatility extends beyond cross-pollination: it can help with self-pollination (possibly replacing bumblebees in controlled situations), gather pollen, and even identify male-sterile or fertile plants by detecting exposed stigmas, eliminating the need for costly DNA testing.
When used with speed breeding (accelerated development through prolonged light cycles) and de novo domestication (rapidly absorbing wild features), GEAIR shortens breeding timelines. The researchers showed this by creating novel tomato lines with richer flavors and greater stress tolerance—the strategy also works for soybeans, pointing to wide agricultural possibilities.
Xu emphasized, “GEAIR isn't just a tool—it's a paradigm shift. We're redesigning crops to unlock AI and robotics, and those technologies are supercharging our ability to create better crops, faster.”
The study highlights a new age in agriculture: Scientists are co-engineering plants and machinery to facilitate quicker, less expensive, and more sustainable crop development. This is crucial as the need for resilient food systems increases globally.
Notably, the method proved transferable. Multiplex gene editing effectively recreated the male-sterile, exserted-stigma phenotype in soybean, a globally important legume. This demonstrates GEAIR's potential application across a wide variety of key crops hampered by similar floral morphology constraints in hybrid breeding.
Journal Reference:
Xie, Y., et al. (2025) Engineering crop flower morphology facilitates robotization of cross-pollination and speed breeding. Cell. doi.org/10.1016/j.cell.2025.07.028.