Beijing, CHINA — For millennia, creating new, resilient crop varieties relied on the delicate dance of pollination – a task nature assigned to bees and wind, and humans painstakingly mimicked by hand. Now, a groundbreaking fusion of biotechnology (BT) and artificial intelligence (AI) has shattered this bottleneck, paving the way for fully automated, high-speed hybrid breeding. Researchers announce the successful development of GEAIR (Genome Editing with Artificial-Intelligence-based Robots), a system that genetically engineers crops for robot compatibility and deploys AI-driven bots to perform precision cross-pollination around the clock.
The Stubborn Bottleneck: Flowers Built for Humans, Not Machines
The quest for higher-yielding, climate-resistant, and flavorful crops hinges on hybrid breeding – crossing distinct parental lines to harness superior traits. Yet, the critical first step, cross-pollination, remained stubbornly manual. Why? The very architecture of key crops like tomatoes and soybeans posed an insurmountable barrier for robots.
"Imagine a tiny, recessed bullseye hidden deep within a complex flower structure," explains Prof. XU Cao, lead researcher on the project. "That's the stigma. Human hands, with their dexterity and vision, could navigate this, albeit slowly and expensively. But for a robot arm, accessing that recessed target without damaging surrounding organs was near impossible." This bottleneck consumed staggering resources: manual pollination accounts for over 25% of fresh-market tomato breeding costs in China alone, with the emasculation step (removing male parts) eating up 40% of that labor. Globally, the economic value of pollination services runs into billions annually.
The Co-Design Breakthrough: Reshaping Flowers for Robotic Hands
Inspired by nature and history – recalling how the Green Revolution bred shorter, sturdier wheat stalks for combine harvesters-the team pioneered a radical "crop-robot co-design" strategy. Their insight: Instead of solely focusing on building better robots, genetically redesign the crop itself to be robot-friendly. Using precise CRISPR-Cas9 genome editing, they targeted conserved "B-class" MADS-box genes in tomato (GLO2) and soybean. Knocking out these genes achieved two crucial transformations simultaneously:
Male Sterility: Preventing the plant from producing viable pollen, eliminating the need for delicate, time-consuming manual emasculation.
Exserted Stigmas: Transforming the once deeply recessed stigma into a prominent, exposed structure---perfectly positioned and sized for robotic access.
"Essentially, we gave the flowers a 'makeover' for the machine age," says Prof. XU. "We reverted a trait artificially selected for inbreeding back to a wild-type form that's ideal for robotic out-crossing."
Enter GEAIR: The AI Cross-Pollinator That Never Sleeps
With the newly engineered male-sterile tomato plants bearing accessible stigmas, the team developed a bespoke mobile robot, the heart of the GEAIR system. Equipped with advanced computer vision powered by deep learning algorithms, the robot roams growth chambers or greenhouses:
Sees: It automatically scans and identifies flowers ready for pollination, specifically recognizing the exserted stigmas.
Acts: A precision robotic arm, guided by AI, gently positions itself and accurately deposits pollen onto the exposed stigma.
The results were striking: GEAIR achieved cross-pollination efficiency comparable to skilled human technicians. But unlike humans, the robot works tirelessly, 24 hours a day, 7 days a week, unaffected by fatigue or the need for breaks.
Beyond Cross-Pollination: A Multipurpose Agricultural Workhorse
The innovation didn't stop there. The researchers realized the core capabilities of the GEAIR robot – advanced visual recognition and precise manipulation – could be adapted for other critical tasks:
Self-Pollination Aid: Equipped with a gentle air-blowing tube, the robot can vibrate flowers to release pollen onto their own stigmas, potentially replacing bumblebees in controlled environments for fruit production.
Automated Pollen Collection: Adding a vibrating rod and collection box allows the robot to harvest pollen from male parent plants efficiently.
Phenotypic Selection Revolution: The exserted stigma is a visible marker for male sterility. GEAIR robots can now autonomously identify and select male-sterile plants (for use as female parents) and fertile plants (for line maintenance or as pollen donors) based purely on visual flower morphology, completely eliminating the need for costly, time-consuming lab-based DNA genotyping. "This shifts breeding from reliance on molecular markers to AI-driven phenotypic markers," emphasizes Dr. XU.
Speed Breeding Synergy and Multi-Crop Potential
Integrating GEAIR with speed breeding techniques – using extended photoperiods to accelerate plant growth cycles – and de novo domestication (rapidly introducing beneficial wild traits into crops) enabled the team to dramatically compress breeding timelines for complex traits. They demonstrated this by rapidly developing new tomato lines with enhanced flavor profiles and improved stress resilience.
Crucially, the strategy proved transferable. Multiplex gene editing successfully recapitulated the male-sterile, exserted-stigma phenotype in soybean, a globally critical legume. This signals GEAIR's potential applicability across a wide range of major crops hindered by similar floral morphology bottlenecks in hybrid breeding.
The Future is Co-Designed: From Green Revolution to Robot Revolution
The GEAIR system represents more than just an automation tool; it heralds a paradigm shift in agricultural innovation. "The Green Revolution adapted crops to machinery. We are now entering an era of 'crop-robot co-design'," states Dr. XU. "We engineer biological traits specifically to unlock the potential of AI and robotics, and in turn, those technologies supercharge our ability to improve crops rapidly."
The researchers envision broader adoption involving robots with multimodal sensors (tactile, olfactory) and more flexible manipulators to tackle even more complex crop traits. The goal: to make the creation of high-performing, climate-resilient, and sustainable crops faster, cheaper, and less reliant on scarce manual labor than ever before.
GEAIR stands as a powerful proof-of-principle: by breaking down disciplinary silos and co-engineering both the plant and the machine, the future of efficient, sustainable, and responsive agriculture is blossoming.
About the Research:
The findings, detailed in the journal Cell on August 11, 2025 under the title "Engineering crop flower morphology facilitates robotization of cross-pollination and speed breeding," result from an interdisciplinary collaboration between the Institute of Genetics and Developmental Biology of the Chinese Academy of Sciences, and partner institutions.
Contact:
Prof. XU Cao
Institute of Genetics and Developmental Biology, Chinese Academy of Sciences
Email: caoxu@genetics.ac.cn
CAS
中文
.png)
