Beijing's ag tech push: bold plans, uneven fields
Pillar Two of the agri powerhouse plan sets the tools—now comes the hard part
The second pillar of Beijing’s agri powerhouse plan doesn’t mince words: technology must lead the way.
But how?
This post unpacks the blueprint and the system behind it—from seed labs in Hainan to IoT greenhouses in Shandong.
Behind the slogans lies a quiet admission: policy tweaks won’t feed the nation.
Might tech?
With rising input costs, a shrinking rural labour force, and slowing yield gains, Beijing is betting on tech to lift agriculture’s productivity ceiling.
Pillar Two of the 2035 Agricultural Strong Country Plan sets out a sweeping innovation agenda — from seed source sovereignty to drone-driven fieldwork — to accelerate self-reliance and close the yield-tech gap.
from labs to land
The 6 May 2025 implementation opinions on boosting agricultural science and technology innovation give shape to Pillar Two of the 2035 Agricultural Strong Country Plan: a high-efficiency system built on tighter coordination, stronger enterprise leadership, and faster commercialisation.
The rhetoric follows familiar lines: ‘new productive forces’, aligned R&D, frontier tech.
Where the 2024–2028 national key R&D areas laid out the what—high-oil soy, low-carbon farming, and the rest—this latest push gets into the how: clarifying institutional roles, rejigging incentive structures, and nudging frontier tech towards something more usable.
New judicial protections for biotech and breeding IP add teeth to what’s otherwise been a fairly toothless system of fragmented capacity-building.
Fan Shenggen 樊胜根, China Agricultural University chair professor, puts it in broader terms: new productive forces (NPFs) are emerging from the crossover of data, biology, materials and energy—layered on top of the old land–labour–capital triad with inputs like knowledge and branding.
The result, at least in theory, is a new engine for modern agriculture.
Getting that engine to turn over is another matter.
Luo Biliang 罗必良, SCAU Agricultural and Rural Policy and Reform Innovation Lab director, breaks it down into three parts
new labour force: most essential, a knowledge-based, skilled workforce who can keep smart kit running and maybe develop a few new practices along the way
new labour tools: everything from robotics and IoT to gene editing, microbiomes, drones, and self-driving tractors
new labour efficiency: big data that actually does something—unlocking new formats like artificial foods, personalised nutrition, and synthetic biology-grown meat, assuming you’ve got the compute and the capital
The 2025 No. 1 Document marks the first official mention of ‘new quality agricultural productive forces’—a phrase that may not trip off the tongue, but signals tech’s upgraded status as the driver of modernisation, noted CNR News.
Talent, meanwhile, remains the real make-or-break.
Wei Houkai 魏后凯, CAAS Rural Development Institute director, argues that training and retaining skilled people in agri-tech R&D, business management, and marketing depends on sorting out the obvious: decent incentives, real career platforms, and somewhere worth living.
He also warns that smallholders—still the dominant production force—can’t be left out of the upgrade. They’ll need to be brought in via benefit- and risk-sharing mechanisms with leading firms, co-ops, and family farms.
Despite a reported 63 percent contribution of sci-tech progress to agricultural growth, structural issues persist.
Lab-firm integration remains patchy, core tech bottlenecks linger, and innovation chains still often break in the usual places, noted Song Jianxiao 宋建晓, deputy director at the CASS Rural Development Institute, writing in Farmer’s Daily April.
Then there’s the hardware.
Much of what passes for ‘smart ag’ still amounts to an ‘intelligence tax’, says Shi Xi 石茜, Sun Yat-sen University professor and AI expert in ag remote sensing. Her 2024 fieldwork across 15 sites found a now-familiar pattern: outdated kit, generic platforms, and giant screens that look impressive and do very little.
Most of the data collected—by her estimate, 99 percent—is unusable. The result is a field full of sensors, a hard drive full of junk, and very little that helps anyone grow anything.
seed source sovereignty
Seed innovation, a cornerstone of Xi Jinping’s storing grain in the land and technology strategy, is critical for ensuring food security and boosting productivity, argues Du Zhixiong 杜志雄 CASS Institute of Rural Development party secretary
breeding bases, policy pushes, and corporate incentives
Han Jun 韩俊, Minister of Agriculture and Rural Affairs, toured the Yanzhou Bay National Laboratory in late 2024—the only PRC state-level agricultural lab.
He called for the lab to act as the ‘main platform and chain leader’ for seed innovation, targeting urgently needed varieties and high-value genes for concentrated breakthroughs.
The broader layout now covers
three national-level breeding bases in Hainan, Sichuan, and Gansu, focused on hybrid corn, rice, cotton, and potatoes
52 national-level breeding counties acting as integration hubs between R&D and production
100 regional specialty bases, covering fruit, veg, nuts, TCM crops, and high-value horticulture
Hainan’s Nanfan Silicon Valley aims to become a global hub for breeding R&D and seed sales by 2030. To lure investment, the provincial government offers
up to C¥10 million for varietal promotion
up to C¥100,000 for technology transformation
a 30 percent subsidy on the annual increase in R&D spending, also capped at C¥100,000
Such moves will launch ‘seed industry revitalisation’ and an internationally competitive seed industry, argued Wu Weihua 武维华 NPC Standing Committee vice chair, at the 2024 China Seed Conference.
The key to unlocking ‘revitalisation’ rest on four pillars, contends Cheng Guoqiang 程国强, National Grain Security Strategy Research Institute director
germplasm protection and use
breeding innovation
sound breeding systems
secure seed supply
regulation, certification, and choke points
Since 2022, MARA has tightened seed approval standards. In December 2024
625 low-quality or duplicate varieties were revoked, a clean up effort to improve IP discipline
new corn, soy, and cotton varieties were approved based on trial yields (e.g. 15,000 kg/ha corn), protein/oil content, and fibre quality
March 2025 saw the official launch of the crop seed certification system, nearly three decades after pilots first began.
The inaugural batch—27 companies, 31 varieties, over 300 batches—was released in Sanya. Germination and purity levels exceeded national benchmarks.
Though codified in the 2016 Seed Law (revised in 2021), full implementation had lagged.
The system is now backed by standardised crop lists, inspection criteria, and certification body accreditation—pitched as both a ‘quality guarantee’ and a ‘market passport’.
But loopholes remain.
A December 2024 Economic Daily op-ed warned that without stricter entry and exit mechanisms, ‘copycat’ varieties will continue to undermine market trust. Recommendations included
shortening DUS (distinctness, uniformity, stability) testing to keep inferior seeds off shelves
improving post-approval tracking and risk assessment, to quickly identify and retire poor performers
structural barriers to innovation
Despite growing policy support, the seed sector still struggles to get promising varieties out of the lab and into the field.
As Bao Xiaoming 鲍晓明, chief scientist at Guorui Biotech, puts it: this isn’t about solving one choke point—it’s about system-level reform.
Issues include
shallow innovation pipelines: even when breakthroughs happen, they often stall before adoption
research–industry disconnect: 88 percent of approved varieties still come from public institutions; fewer than 100 of over 7,000 seed firms have active R&D capacity
short-cycle funding: project-based models limit multi-year varietal development
data fragmentation: phenotype datasets—critical for AI-assisted breeding—are siloed across platforms, slowing refinement
To tackle the financing gap, MARA and the Agricultural Development Bank launched the Seed Industry Revitalisation Loan in January 2025. It offers fast-tracked financing tied to research–commercialisation pipelines.
But tech isn’t the whole answer.
Since Reform and Opening-up, over 6,500 vegetable varieties have been developed, progressing through four to five generations of seed upgrades.
As Zhang Yangyong 张扬勇, CAAS Institute of Vegetables and Flowers deputy director, notes
despite progress in veg variety germplasm collection, use innovation, and genomic research, this ‘golden age of development’ still lags ten years behind developed countries
As it stands, PRC is still in ‘breeding 3.0’ (marker-assisted selection), while leading countries move into ‘breeding 4.0’: gene editing, whole-genome selection, and smart design.
Major constraints remain in
functional gene analysis and trait-targeted breeding
investment in infrastructure and core breeding tech
low levels of basic-to-cutting-edge research integration
weak original innovation capacity
The bottom line: self-reliance won’t be won on regulation alone. It will take a wholesale rethink of how research, finance, and commercialisation connect—plus stronger systems for tracking what works, and what doesn’t.
machinery modernisation
Mechanisation is a cornerstone of the agri-tech self-reliance push. The strategy blends fiscal muscle, industrial policy, and technical ambition—but the sector remains stuck between fragmentation, uneven coverage, and chronic shortfalls in high-end capability.
targeted subsidies and fiscal support
MARA and MoF issued the implementation opinions on agricultural machinery purchase and application subsidies (2024–26) on 4 June 2024.
The goal: foster ‘new quality productive forces’ by subsidising equipment that
boosts staple crop yields (e.g. high-speed transplanters, precision seeders, smart harvesters)
works in hilly and mountainous areas
fills equipment gaps or improves weak links
supports emerging ag subsectors
With little room left to expand farmland, mechanisation and crop optimisation are now the primary levers for yield growth, notes Li Guoxiang 李国祥, CASS researcher.
The 2025 No. 1 Document echoes this—flagging precision ag, high-standard farmland, and stricter land-use discipline as key.
Fiscal policy is doing the heavy lifting. Beijing is raising the deficit ratio, issuing ultra-long-term special bonds, and boosting local debt quotas. These funds back everything from machinery upgrades to rural infrastructure.
At the same time, monetary easing is keeping liquidity flowing. With ag machinery finance needs up over 30 percent in 2024, low rates are crucial to keeping the capital-intensive sector afloat.
structural mismatches and uneven progress
Total ag machinery power exceeded 1.078 billion kW by 2023 YE, with 73 percent of crop cultivation mechanised. But
cash crops, livestock, fisheries, and processing remain under-mechanised
smart equipment uptake is limited
integration with agronomic practices is weak
core components still face technical bottlenecks
Mechanisation passed 75 percent in 2024, hitting the 14FYP target a year early, MARA confirmed at April’s 2025 National Mechanisation Work Conference.
But progress is lopsided.
The sector still follows the familiar ‘three highs, three lows’ pattern:
high mechanisation in staples, low in cash crops
high in northern plains, low in southern hills
high in planting, low in processing and livestock
Market segmentation mirrors this divide.
top-tier firms offer strong brands and terrain-adapted products
second-tier firms have broad lines but lag on innovation
third-tier firms often sell over C¥1 bn, but lack R&D
exports up, demand flat, quality gaps persist
Despite soft domestic demand, ag machinery exports rose ~12 percent in 2024, outperforming overall trade.
This reflects improved product quality, stronger supply chains, and rising internationalisation, argues nongjitong.com (an ag machinery industry information website). Export structures are shifting away from hand-held models toward 50–180 HP tractors.
But the upgrade is shallow.
PRC firms still lag in core tech and high-end patents, especially in
power-shift and CVT tractors
large beet and tomato harvesters
advanced electronic control systems
Field-readiness is another issue.
Many Q1 2025 releases had not yet been tested in complex conditions.
Overcapacity coexists with gaps in high-performance, precision models.
performance over price
Buyer expectations are shifting. According to partial data
20,000 medium–large tractors sold — flat y-o-y
200+ HP and power-shift models hit record highs
wheat harvester sales neared 10,000 units, with 10–12 kg/s intake and vertical axial-flow types leading growth
Buyers are demanding more from their machines. But low grain and livestock prices have squeezed farm margins and purchase intent is down.
Even with raised subsidies for hilly terrain equipment, larger farms remain best positioned to benefit.
Smallholders, once again, risk being left behind.
digital agriculture’s long road
Smart agriculture is central to the agri-modernisation plan. Beijing is pushing ahead with funding and frameworks, but bottlenecks in infrastructure, skills, and adaptability still hold back meaningful gains.
policy blueprints and pilot zones
The 2024–2028 smart ag implementation plan and supporting guiding opinions set the direction. Their priorities
build national-level smart ag demo zones
create a unified national ag data platform
support digitalisation across production, logistics, and sales
These efforts aim to boost productivity, raise farmer incomes, and improve sustainability.
Farmers Daily warns that climate change is already destabilising major growing regions.
For Beijing, that only sharpens the urgency behind the digital shift.
Fuxi Farm and the replicable smart ag model
At the 30-ha Fuxi Farm in Shilipu village, Xiong’an New Area, the Chinese Academy of Sciences Xiongan Innovation Institute is testing a scalable smart ag model.
Applications include
precision soil analysis
traditional satellite or sample-based testing lacks accuracy
the Gamma Spectrometer delivers square-metre precision, feeding a detailed soil database
autonomous machinery
driverless tractors use sensors, LiDAR, and Beidou navigation to perform tasks like corn planting without human input
digital twin simulation
virtual crop models forecast growth, pest threats, and climate stress
simulations help optimise inputs and reduce risk
AI-assisted breeding
large datasets identify genetic markers linked to target traits
AI models simulate plant growth under varied conditions, compressing breeding cycles from years to days
real-time data monitoring
The farm’s command centre tracks soil moisture, land use, and weather
In theory, this guides precise, data-driven decisions on planting and irrigation
The model is already expanding.
In Hubei, a 40-ha smart farm is under construction—tailored to fragmented, hilly plots.
In Shouguang, Shandong, 85 percent of new greenhouses are IoT-enabled, automating temperature, humidity, and irrigation.
barriers to adoption and scale
Yet bottlenecks remain.
116 smart ag pilots were launched and seven key recommended technologies released by 2024 YE. But
digitalisation in primary production remains low (10.78 percent penetration rate)
core tech like sensors and chips still face bottlenecks
Economic Daily highlights familiar issues
insufficient tech support and high upfront costs
trend-chasing and redundant image projects
lack of local adaptation, making systems hard to scale
weak digital skills among farmers
MARA proposes a multi-pronged solution
fiscal subsidies and tax incentives
tailored infrastructure upgrades
greater private-sector participation
region-specific support and stronger supervision to avoid wasteful megaprojects
Regular demo cases will be published to provide replicable models across regions.
innovation systems and the talent gap
MARA stresses the need to link digital rollout with innovation systems. The plan calls for
tiered, cooperative R&D networks balancing competition and collaboration
coordination between universities, firms, and research institutes
targeted research in crop genetics, machinery, pesticides, soil quality, and water-saving tech
improved IP systems to reduce duplication and strengthen knowledge transfer
Talent remains a critical constraint. The plan puts rural human capital front and centre—calling for:
digital literacy and management training for farmers
a professionalised, tech-savvy rural workforce
political incentives to attract and retain skilled individuals in the countryside
The ultimate goal is to make farmers—not just tech platforms—the drivers of smart ag.
from system to soil
Smart ag is often cast as the solution to the land–labour mismatch and the key to high-quality development.
But as Shi Xi 石茜 warns, that success depends on closing the gap between innovation and practice.
Agri tech must deliver real value for crops and farmers—not just tick policy boxes.
from tools to table
If Pillar Two builds the tools, Pillar Three sets the table—with rural business reform, collective economies, and a push to modernise China’s fragmented producer landscape.
More on that next week.