 |
| Next to drones and IOT detectors, China uses satellites to feed AI systems to optimise agricultural cultivation |
How is AI helping China achieve its food security goals?
AI is a cornerstone of China's agricultural modernisation strategy, specifically designed to address the challenges of a large population and limited arable land while ensuring national food security. By maximising the efficiency of existing farmland, AI allows China to pursue self-sufficiency through precision management and technological innovation.
Precision Resource Management: AI and big data analytics enable precision farming, which optimises inputs like water, fertiliser, and pesticides to boost yields and reduce waste. For instance, digitised greenhouses in Shouguang use AI-monitored systems to save 40% of water and 30% of fertiliser while increasing vegetable production by roughly 5%.
Mitigating Labour Shortages: As the rural workforce ages and migrates to cities, AI-powered autonomous robots and drones fill the labour gap. Drones are used for precision spraying and seeding, while specialised robots now perform delicate tasks such as picking tea leaves and tapping rubber trees.
Pest and Disease Control: AI systems, such as the "Tian-zhi-1" project, combine satellite remote sensing with AI to predict the spread of pests and diseases. This allows for proactive prevention, as demonstrated when the system successfully forecast a rice planthopper outbreak in 2019, thereby safeguarding food supplies.
Smart Breeding and Genetics: AI-assisted gene editing and big data analysis allow scientists to identify key genes for high-yield, drought-resistant, and disease-resistant crop varieties. Some AI-driven breeding projects have reported speeding up crop development by up to 400%.
Vertical Urban Farming: To overcome land constraints, China is pioneering vertical "plant factories" and "farm scrapers". The 51-story Jian Mu Tower in Shenzhen, for example, integrates a vertical hydroponic farm designed to produce enough food to support 40,000 people annually.
Livestock and Aquaculture Monitoring: AI cameras and sensors monitor the health, weight, and behaviour of livestock and fish to prevent disease outbreaks and optimise feeding. This ensures a stable supply of animal protein while reducing economic losses from mass mortality.
These efforts are supported by a robust national policy framework, including the Smart Agriculture Action Plan, which aims for more than 30% of agricultural production to be led by information-driven systems by 2026. Infrastructure investments, such as 5G-enabled unmanned farms, provide the high-speed connectivity needed for real-time remote monitoring and automated machine operation. Acknowledging the role of private industry, the government also encourages social capital investment and partnerships with tech giants like Alibaba, JD.com, and Huawei to accelerate these breakthroughs.
Information about the 51-story vertical farm in Shenzhen.
The 51-story vertical farm in Shenzhen, known as the Jian Mu Tower, is designed to be the world's first "farm scraper," a 218-meter-high building that integrates large-scale urban agriculture directly into its structure. Located on the final available plot in Shenzhen's Central Business District, the tower merges office spaces, a supermarket, and a food court with a vertical hydroponic farm that extends the entire height of the building's façade.
Key features and capabilities of the Jian Mu Tower include:
Food Production and Sustainability: The tower features 10,000 square meters of space on its façade dedicated to crop cultivation. It is estimated to produce roughly 270,000 kilograms of food per year, which is enough to support the needs of approximately 40,000 people. This design creates a self-sustained food supply chain where crops are grown, harvested, sold, and consumed all within the same building.
AI-Managed Farming: The farm's daily operations, including managing irrigation and nutritional conditions for crops such as salad greens, fruits, and aromatic herbs, are overseen by an AI-supported "virtual agronomist".
Environmental Benefits: Beyond food production, the vertical farm provides solar shading, which reduces indoor solar irradiance and the subsequent need for air conditioning and a greater water supply. The building also utilises a sustainable irrigation system that takes advantage of Shenzhen’s abundant precipitation to water a variety of flora on landscape terraces, including water lilies, ferns, and lychees.
Design and Symbolism: Designed by Carlo Ratti Associati (CRA), the tower's name is inspired by the mythical "jian mu tree," which in ancient Chinese folklore connects heaven and earth. Its architectural form mirrors this symbolism, featuring a rectangular base that gradually morphs into a tubular form as it rises.
Tenant Experience: The tower offers 90,000 square meters of space and integrates double-height gardens into its interior to immerse office workers in a natural setting. Additionally, tenants can use a mobile app to customise the micro-climatic conditions of their specific office spaces.
The project was envisioned for an international competition organised by the Chinese supermarket chain Wumart and involves collaborations with companies like ZERO, which specialises in innovative hydroponic farming systems. So, instead of supplying false information to consumers – as often is the case in Europe and the US, this is the proper way to inform them correctly. This probably stems from the awareness in the Chinese government that its labour force must be healthy and intelligent.
mountain into farm land
How does the virtual agronomist manage the crops?
Humanoid robots like Xunxiao navigate the complex and unstructured environments of tea gardens using a sophisticated navigation system based on high-precision 3D fusion technology. This technology allows the robot to create real-time three-dimensional maps of its surroundings, enabling it to identify and negotiate obstacles such as stairs, slopes, and perforated surfaces.
Key technical aspects of its navigation and operation include:
Precision Positioning: Even in crowded or complex spaces, Xunxiao maintains positioning accuracy within 0.5 centimetres.
Embodied Intelligence: Developed by the Youibot-XJTU research institute, the robot combines vision-language models with specialised expert systems to determine pest and ailments in an early stage. This allows it to handle the "long task chains" and flexible operations required in non-structured environments that traditional industrial robots cannot manage.
Precise Interaction: This navigation and vision system enables the robot to move delicately through gardens, using its robotic arms to select and pluck only the freshest tea buds with surprising precision.
Beyond tea gardens, this advanced navigation and "embodied intelligence" framework allows the robot to be repurposed for other complex tasks, such as inspecting power equipment or warehouse sorting, without requiring hardware modifications.
plowing a farm field in China
What other tasks can the Xunxiao robot perform besides tea picking?
What other tasks can the Xunxiao robot perform besides tea picking?
The Xunxiao humanoid robot is designed as a versatile "embodied intelligence" system that can be deployed for various complex assignments beyond tea picking without requiring hardware modifications when switching between tasks.
According to the sources, the Xunxiao robot is capable of performing the following additional functions:
Power Equipment Inspection: The robot can also be used to inspect infrastructure in the energy sector, that is part of a larger system to balance energy consumption between energy sources in that system.
Semiconductor Precision Operations: It is capable of handling delicate tasks within the high-tech manufacturing of specialised semiconductors.
Intelligent Warehouse Sorting: The robot can manage logistical duties such as sorting items in warehouses and determine the time to harvest.
These
capabilities are made possible by the robot's combination of
vision-language models and specialised expert systems, which
allow it to handle "long task chains" and flexible
operations in non-structured environments. Its advanced navigation
system also supports these tasks by creating real-time 3D maps
and maintaining positioning accuracy within 0.5 centimetres,
even in crowded or complex settings.
rubber tapping robot in China
In the southern island province of Hainan, smart rubber-tapping robots solve labour shortages by automating the physically demanding and precise task of harvesting rubber latex. These self-navigating robots are designed to move through plantations independently, identifying individual rubber trees and approaching them with pinpoint accuracy, long before humans detect that intervention is necessary. Once in position, the robot extends its robotic arm to execute a precise cutting motion on the tree trunk to facilitate the flow of rubber.
This technology is a critical component of China's broader agricultural strategy for several reasons:
Priority Innovation: The Chinese government has officially identified rubber-tapping automation as a major priority for high-quality AI applications in the economic sector.
Filling the Labour Gap: These robots serve as a sought-after alternative in areas facing significant shortages of skilled manual workers.
Embodied Intelligence: These systems represent a shift toward 'embodied artificial intelligence', which combines advanced AI models with physical robotic capabilities to handle complex, unstructured tasks that traditional machinery cannot manage.
Precision and Efficiency: By using high-precision technology, these robots can operate with a level of accuracy that ensures the health of the tree in order to maintain production efficiency.
How do robots accurately identify rubber trees and cut them?
Rubber-tapping robots in China, particularly those deployed in the southern island province of Hainan, achieve accurate and timely identification and determine harvesting time through a combination of autonomous navigation and high-precision mechanical control.
The process works as follows:
Autonomous Identification and Navigation: These robots are self-navigating systems that move through plantations independently. While the sources do not explicitly detail the internal sensors of the specific rubber-tapping model, they highlight that this new generation of versatile agricultural robots utilises high-precision 3D fusion technology to create real-time three-dimensional maps of complex, unstructured environments. This allows the system operators to approach individual rubber trees and assess with pinpoint accuracy if action should be undertaken.
Precise Cutting Execution: Once the robot is correctly positioned at the tree, it extends a specialised robotic arm to execute a precise cutting motion on the trunk. This automation is designed to handle the delicate task of facilitating rubber latex flow, a job that traditionally requires skilled manual labour.
Embodied Artificial Intelligence: These capabilities are part of China’s push into 'embodied intelligence,' which combines vision-language models with specialised expert systems. This technology enables robots to manage 'long task chains'
and flexible operations in environments that traditional industrial robots find too complex to navigate.
The
development of these robots is a strategic priority; the Chinese
government has officially identified rubber-tapping automation
as a major goal for high-quality AI applications to solve labour
shortages and modernise the economic sector, while elevating production. Controlled cultivation results in water and energy saving, while preventing the spread of pests and ailments.
How does AI-assisted gene editing speed up crop development?
AI systems predict the spread of crop pests by integrating multi-source data collection with advanced analytics to create early warning models - detecting leaf and root colours and soil moisture levels. These systems move beyond simple detection to forecast outbreaks based on environmental trends and biological patterns.
According to the sources, the prediction process involves several key technological layers:
1. Data Collection and Remote Sensing
Satellite and Remote Sensing: Projects such as "Tian-zhi-1" combine satellite remote sensing with AI to monitor vast areas and predict how pests and diseases will move across regions.
Drones and High-Resolution Imagery: Drones capture high-resolution horticultural maps and real-time data on pest distribution, providing the visual input necessary for AI models to assess current threats.
IoT and Ground Sensors: Field-deployed sensors monitor environmental factors that influence pest life cycles, such as temperature, rainfall, soil moisture, and mineral levels.
Specialised Robotic Monitoring: In regions like Sichuan, autonomous robots inspect fields using black light cameras and built-in AI to flag pests that might not be easily visible to the human naked eye.
2. Analytical Models and Forecasting
Big Data and Machine Learning: Platforms like JD Farm utilize machine learning and big data analytics to process collected environmental data to continuously guard cultivation efficiency. These systems analyse historical patterns and current trends to forecast crop growth and potential pest outbreaks.
Integrated Platforms: Systems such as Microsoft's Azure Data Manager for Agriculture or Alibaba's ET Agricultural Brain integrate diverse data—including satellite imagery, weather forecasts, and soil data—into a single model to support intelligent decision-making and predictive alerts.
'Virtual Agronomists': In urban vertical farms like the Jian Mu Tower, AI-supported 'virtual agronomists' manage the day-to-day operations, including monitoring for conditions that could lead to pest issues.
3. Early Warning and Proactive Prevention
The primary goal of these systems is to shift from reactive treatment to proactive prevention.
Outbreak Prediction: In 2019, the "Tian-zhi-1" system successfully predicted an outbreak of rice planthopper in Yunnan Province, allowing farmers to take preventive measures before significant damage occurred.
Early Alert Systems: Based on AI analysis, farms can deploy early warning systems that alert farmers to impending risks, helping to minimise crop losses and reduce the overall volume of pesticides needed by targeting only high-risk areas.
Adapting to Shifting Patterns: Research initiatives, such as those by Google's Alphabet X, focus on helping farmers adapt their practices to shifting pest patterns caused by climate change.
Information about the Tian-zhi-1 project.
The Tian-zhi-1 project is a pioneering Chinese initiative that combines satellite remote sensing with artificial intelligence to enhance agricultural monitoring and food security. It is designed to provide high-performance, on-orbit intelligent computing and real-time services, utilising space-based large-scale computing power to analyse agricultural data.
Key features and achievements of the project include:
Pest and Disease Prediction: The primary application of Tian-zhi-1 is to predict the spread of crop pests and diseases. By analysing satellite data with AI, the system identifies potential outbreaks before they happen, allowing farmers to take proactive, preventive measures rather than simply reacting to damage.
Success in Yunnan Province: In 2019, the project demonstrated its practical value by successfully forecasting an outbreak of rice planthopper in Yunnan Province. This early warning was instrumental in reducing crop losses and safeguarding the local food supply.
Technological Sophistication: The project represents a shift toward intelligent remote sensing, where data processing happens on the satellite itself (on-orbit) to provide faster, more actionable insights for farmers and decision-makers.
Impact on Food Security: By providing a technological "early warning system," Tian-zhi-1 helps minimise the environmental and economic impact of pests, contributing to China's broader goal of agricultural modernisation and self-sufficiency.
How does the Tian-zhi-1 project use intelligent remote sensing?
The Tian-zhi-1 project utilises intelligent remote sensing by integrating satellite data with artificial intelligence to predict the spread of crop pests and diseases. This system represents a technological shift toward on-orbit intelligent computing, which allows for high-performance processing directly in space.
Key aspects of how the project uses this technology include:
Real-Time Actionable Insights: By leveraging space-based, large-scale computing power, the project provides real-time services. Processing data on-orbit rather than sending all raw data to Earth first allows for faster delivery of critical information detection to farmers.
Proactive Disease and Pest Management: The system moves beyond simple detection to forecast outbreaks before they happen. This capability was demonstrated in 2019 when it successfully predicted a rice planthopper outbreak in Yunnan Province, allowing farmers to implement preventive measures that safeguarded food security.
Support for Agricultural Modernisation: As part of China's broader policy to integrate AI into various industries, Tian-zhi-1 serves as a foundational example of integrated remote sensing networks used for agricultural monitoring and data-driven decision-making.
By providing this technological "early warning system," the project helps minimise environmental and economic impacts while contributing to national food security goals.
Information about JD Farm's AI agricultural practices
JD Farm, an initiative by the e-commerce giant JD Group, leverages big data, the Internet of Things (IoT), and AI to drive the digital transformation of Chinese agriculture. Centred on the philosophy of 'ecological agriculture, healthy dining,' the project focuses on establishing modern agricultural bases through scientific planting, standardised production, and efficient logistics.
JD Farm’s practices are categorised into four primary areas:
1. Precision Farming via Digital Modelling
JD Farm creates a digital model of farmland to manage the entire lifecycle from planting to harvest.
Remote Sensing: Satellites, drones and IOT detectors capture high-resolution imagery to provide real-time data on crop growth, soil moisture, and pest distribution.
Soil Monitoring: Sensors monitor temperature, moisture, and nutrient content, providing the scientific basis for accurate fertilisation and irrigation.
Smart Machinery: This data integrates with intelligent machinery to perform precise planting and variable-rate fertilisation, which maximises resource use and promotes sustainability.
2. Intelligent Management and Forecasting
The farm utilises a comprehensive data network to monitor and control crop environments in real time.
IoT Sensor Networks: Sensors installed in fields collect environmental data such as light, humidity, temperature, and carbon dioxide levels.
Predictive Analytics: Using machine learning and big data, JD Farm analyses this data to predict pest and disease outbreaks and track growth trends.
Early Warning & Automation: These systems provide farmers with alerts to take preventive actions. Additionally, intelligent control systems can automatically adjust conditions in greenhouses to optimise production.
3. Efficient Operations and Traceability
JD Farm integrates technology across the entire supply chain to ensure quality and efficiency.
Blockchain Traceability: Blockchain records information at every stage—planting, processing, and transport—to ensure data transparency and immutability. Consumers can scan QR codes on products to access this full history.
AI Logistics: JD’s logistics network uses AI algorithms to optimize delivery routes, ensuring products reach consumers quickly while minimising transport losses.
Market Analysis: Big data is used to analyse market demand, which helps optimise inventory management and production planning.
4. Livestock Monitoring
Beyond
crop production, JD Agriculture (part of JD's broader
agricultural efforts) applies AI to animal husbandry.
Chinese customers check
food history with QR code
food history with QR code
What data can customers see by scanning the QR codes?
AI Cameras: In pig farms, AI-powered cameras monitor animal weight, behaviour, and health and Disease Control: The system automatically alerts staff if abnormal behaviour is detected, allowing for timely medical intervention and preventing the mass spread of disease.
How does blockchain ensure traceability in JD Farm's products?
Blockchain ensures traceability in JD Farm's products by creating a transparent and unalterable record of a product’s journey throughout the entire supply chain, from 'farm to table'.
The system achieves this through the following mechanisms:
Comprehensive Recording: Blockchain is applied across every stage of the lifecycle, including planting, processing, transportation, and sales. Data is recorded at each step to document the product's history.
Data Immutability and Transparency: A core feature of blockchain is that it ensures data transparency and immutability. Once information is entered into the system, it cannot be changed or falsified, which guarantees the integrity of the data regarding the product's origin and handling.
Consumer Access via QR Codes: Consumers can verify this information by scanning QR codes on the product packaging. This allows them to access the full history of the item - such as the specific planting site, the amount of pesticides used, and the conditions during transport - which significantly enhances customer trust in the quality and safety of the food.
By integrating blockchain with other technologies like AI and the Internet of Things (IoT), JD Farm facilitates a more transparent and precise decision-making process for consumers, suppliers, and business partners alike.
How does JD Farm use AI to monitor livestock health and optimise delivery routes?
By scanning QR codes on agricultural products, customers can access detailed traceability and quality data intended to verify food safety and build consumer trust.
According to the sources, the specific data available to consumers includes:
Origin and Planting Site: Information detailing exactly which farm or planting site the product came from.
Chemical Usage: Records of how much pesticide was used during the cultivation process.
Full Production Lifecycle: The ability to trace the entire process from the initial breeding and planting stages through processing and transportation.
Quality and Maturity Assessments: Data regarding the safety, quality, and maturity of the produce, often verified by machine vision systems.
This
system is frequently powered by blockchain technology, which
ensures that the recorded information is transparent and cannot be
altered, providing a reliable "farm to table" history for
the consumer.
What is the Smart Agriculture Action Plan (2024–2028)?
The Smart Agriculture Action Plan (2024–2028) is a strategic initiative introduced in October 2024 by China's Ministry of Agriculture and Rural Affairs (MARA) to modernise the nation's agricultural sector through digital and intelligent technologies,. It serves as a coordinated policy framework designed to leverage existing projects and resources to accelerate the transition from traditional to smart farming.
The plan outlines several key objectives and implementation strategies:
Core Objectives
Intelligent Transformation: The plan seeks to promote the intelligent transformation of the entire agricultural production chain, moving beyond isolated technologies to integrated, data-driven systems.
Advanced Robotics and Machinery: A primary focus is advancing the practical application of smart agricultural machinery and agricultural robots. This includes technologies for specialised tasks like orchard weeding, precision crop protection, and livestock management.
Demonstration Zones: The initiative aims to establish smart agriculture demonstration zones to test and showcase high-efficiency farming models.
Implementation and Support
Infrastructure and Funding: To improve smart agriculture infrastructure, the plan aims to integrate various funding channels and encourage the participation of social capital.
Technological Breakthroughs: The plan prioritises accelerating breakthroughs in critical areas, including the development of agricultural sensors, specialised chips, core algorithms, and large-scale AI models.
Information-Driven Target: Under this broader national strategy, the government expects more than 30% of its agricultural production to be led by information-driven systems by 2026.
This action plan is supported by a network of 34 innovation labs and 35 IT laboratories dedicated to smart agriculture, spearheaded by the Chinese Academy of Sciences to ensure groundbreaking advancements are fostered across various provinces.
