Plant agriculture
Plant agriculture covers various areas of technologies related to open field crop agriculture based on plants and trees. They span the entire crop-production cycle ranging from the working of the soil i.e. by tillage (ploughing), seeding and fertilizing to harvesting the crops and preventing spoilage during their storage. Forestry also falls into this category. In each of these five areas, digitalization creates significant benefits by enabling higher precision - even up to individual plant level - that can increase yields while minimizing resources and inputs such as water, fertilizers, seeds and labour.
- Soil working
Soil working aims to improve soil structure, enhance aeration and increase water infiltration; ultimately creating optimal conditions for seed germination and plant growth. Soil working covers all relevant details on tools, methods or agricultural machinery parts used for an action that will result in more precise furrowing, ploughing, tilling, moving, opening, smoothing, etc. Nowadays, digital tools and technologies are transforming soil working practices, making them more sustainable and efficient, and helping farmers to optimize crop production while minimizing its environmental impact.
Sensing and imaging
Soil sensors and imaging tools provide detailed maps of moisture, compaction and texture that guide targeted tillage for an ideal seedbed. For example, moisture probes mounted on a tractor toolbar can relay real-time soil‐water readings, prompting the operator to postpone ploughing overly wet patches and avoid unnecessary field passes.
Data processing and artificial intelligence (AI)
By feeding historical yields, soil tests and weather data into machine-learning models, farmers receive precise recommendations on, for example, where and how intensely to till. A farm management platform may analyse past compaction patterns and steer GPS-guided subsoiling on just a part of the field, cutting fuel use and labour while improving the conditions for optimal root penetration.
Automation and robotics
Autonomous tillage machines use advanced GPS guidance and implement sensors to automatically adjust depth and degree of soil disturbance, ensuring consistent and optimal soil preparation without driver fatigue. It also helps to retain soil moisture.
Drones
Multispectral drones reveal crop-stress patterns linked to subsoil compaction, enabling spot aeration rather than whole-field tillage—thermal imagery, for instance, can highlight stunted rows where targeted subsoiling will restore root growth.
- Seeding and fertilising
Seeding involves planting seeds, seedlings, plants, bulbs and/or tubers in the soil to grow crops. Fertilising involves adding nutrients to the soil to enhance plant growth and yield. Fertilisers can be organic (e.g. compost, manure) or inorganic (e.g. chemical fertilisers). High-precision autonomous seeding and fertilising can be based on AI and machine learning to analyse images of the seeding process and crops to establish their current state and estimate fertilisation needs. Historical data can be used to recommend precise planting patterns. Precision fertilising can substantially reduce the total amount of fertiliser used by a farmer. Nitrogen reductions of up to 75% have been reported using variable-rate nitrogen application, without affecting product yield.
Sensing and imaging
High-quality sensors yield more accurate and reliable data, which are fundamental for effective AI-based applications in digital agriculture. They can provide critical information for seeding and fertilizing. During planting season, for example, drills fitted with real-time optical sensors and flow meters may real-time verify that seeds and fertilizer are being deposited at the correct rates and depths. Cameras can monitor seed furrow closure and seed-to-soil contact, sending alerts if blockages or skips occur. Meanwhile, near-infrared reflectance sensors can gauge soil nutrient levels just ahead of the seeder. By combining these information streams, operators can immediately adjust settings to ensure uniform plant populations and optimal nutrient placement.
Data processing and AI
AI-driven prescription maps can blend soil fertility, yield history and forecast data to vary seeding rates and fertilizer doses. AI-powered farm-management platforms, for example, can merge soil-test data, historical yield variability maps and local weather forecasts to create variable-rate application (VRA) prescriptions. These prescriptions detail the exact quantity of seed and fertilizer needed at each GPS-defined grid cell. Once uploaded to VRA planters, the system can then dynamically modulate seed metering and fertilizer valves on the go, matching the agronomic potential of each zone.
Automation and robotics
Using data from sensors, GPS and AI input, agricultural machinery can perform precision seeding, planting and fertilizing. Robotic planters, for example, can autonomously modulate seed and meter out fertilizer based on GPS position and soil feedback. They can also slow application on sandy patches, for instance, to maintain consistent seed depth without operator input. Precision fertilising can substantially reduce the total amount of fertiliser used and minimise runoff into water bodies.
Drones
In-season drone operations supplement ground-based planting by delivering micronutrient sprays or targeted foliar feeds to young seedlings. After reviewing post-emergence plant quality maps, drones can be sent to hover over deficient patches and apply liquid supplements via precision nozzles. This aerial foliar feeding corrects nutrient gaps before they stunt early growth and reduces the need for broad-spectrum ground sprayers. It is especially effective in fields where tractor access is limited by wet conditions or narrow terraces.
Minimum till seeding
Minimum till seeding means reduced or conservation tillage. It is a farming practice that minimizes soil disturbance during seedbed preparation, with a view to reducing erosion and improving soil health. It involves using shallower cultivations than traditional methods like ploughing and often involves techniques like strip tillage or using tines or discs to create narrow planting rows.
Minimising gas pollutants and runoffs
Practices that reduce nitrogen loss and enhance nutrient uptake can minimize the emission of gas pollutants during seeding and fertilizing. They include using the right fertilizer type, applying it at the correct time and rate, and even considering alternative fertilization methods like biological nitrogen fixation or organic fertilizers. Practices such as no-till farming and improved manure management can also significantly reduce emissions and runoff.
- Harvesting
Harvesting covers a wide range of technologies and methods related to the harvesting of crops and mowing of grass or similar vegetation. The invention of the combine harvester in the nineteenth century revolutionised grain harvesting by combining reaping, threshing and winnowing into a single operation. Today, digital agriculture is transforming harvesting by utilising technologies like GPS, sensors and data analytics involving AI to enable precision harvesting, optimise timing and improve efficiency and sustainability in crop collection.
Sensing and imaging
During harvest, combines can be outfitted with light detection and ranging (LiDAR) and hyperspectral sensors that scan standing crops to determine moisture content, kernel size distribution and lodging areas in real time. These sensors feed data to the harvester’s control unit, which subsequently adjusts reel speed, header height and threshing cylinder settings to optimize grain separation and minimize shatter losses. The integrated imagery also produces yield-quality maps that later guide storage allocation and post-harvest handling. This smart sensing helps to ensure that each harvested load meets quality standards, reducing dockage and maximizing market value.
Data processing and AI
The use of data and AI can impact harvesting in many ways. For example, machine learning algorithms can predict optimal harvest timing by analysing crop maturity, weather forecasts and soil conditions. AI models can recommend optimal harvest routes to minimize travel time between full bins, advise on when to switch unloading locations and even forecast end-of-day tonnage for logistics planning such as the storing and transport of harvests. This real-time decision support empowers operators to adapt to changing field conditions, like wet spots or crop displacement, without returning to base. Cloud platforms store and analyse yield data from IoT-enabled harvesters enabling farmers to access real-time updates on harvest progress and machine performance via cloud systems.
Automation and robotics
Autonomous harvesters, guided by high-precision positioning, carry out end-to-end grain harvesting with minimal human oversight. From navigation and crop cutting to threshing and grain tank unloading, every process is automated based on preset algorithms and continuous sensor feedback. The machines self-diagnose performance, alert support staff to maintenance needs and roam fields 24/7 under operator supervision in a remote-control centre. This automation addresses labour shortages during peak harvest windows and ensures consistent operation, regardless of human fatigue.
Drones
Preharvest drone surveys provide aerial intelligence on crop maturity uniformity, moisture gradients and lodged patches. By flying grid patterns at dusk (when canopy temperature differences are pronounced), thermal imaging drones pinpoint areas that have dried sufficiently for safe combine entry. Operators then prioritize these zones to reduce kernel cracking or spoilage. Drones may also provide real-time sensor-support during harvesting operations.
- Spoil reduction
In agriculture, spoil reduction involves strategies and practices aimed at minimizing the loss of crops and food products across the supply chain, from production to consumption. This includes improving harvesting techniques, enhancing storage conditions, optimizing transportation, as well as implementing efficient processing and packaging methods to extend shelf life and maintain quality, thereby reducing waste and increasing food availability. This encompasses technical aspects such as machines and methods used to separate grains from their husks or straw, devices for handling harvested crops such as conveyors, elevators and other machinery, and logistics improvements to optimize storage and transportation. The use of AI algorithms can optimize storage and transportation routes to reduce delays and food waste. Cloud platforms can track the quantities and quality of produce in real-time, improving supply chain and energy efficiency.
Sensing and imaging
Machine vision systems can classify produce by size, quality and ripeness so that top quality products end up on the shelf and lesser quality products are used in industrial food production processes. It is equally important to monitor moisture levels during storage and production, which can lead to economic losses due to germination and mould development.
Data processing and AI
Cloud platforms can aggregate storage-facility sensor data with external weather forecasts to run spoilage-risk models. AI algorithms calculate optimal aeration schedules, humidity setpoints and the duration of fan operation to maintain target moisture levels. The system sends automated commands to ventilation controls or alerts operators when manual intervention (e.g. grain turning) is necessary. AI can also help with inventory management and with planning transport logistics to reduce time to market.
- Forestry
Forestry in agriculture, often referred to as agroforestry, includes managing forest resources for timber, non-timber products and ecosystem services. It covers various aspects of forestry, including the cultivation, management and conservation of forests, technologies and methods for planting, maintaining and harvesting trees, as well as managing forest ecosystems for environmental benefits. Forestry is a complex and multifaceted field that balances ecological, economic and social considerations, highlighting the need for sustainable management practices. Mapping and precision agroforestry are enhanced by the use of drones and AI algorithms.
Sensing and imaging
Forestry management may employ ground and aerial LiDAR to construct 3D models of stand structure, capturing tree height, canopy density and understory biomass. Combined with multispectral imagery, these maps reveal growth rates, drought stress and pest-infested patches. Forest managers use this data to prioritize thinning operations, ensuring resources go to stands with the greatest yield or conservation value. The imagery also documents changes over time, helping to evaluate the needs for thinning, harvesting, planting, pruning, prescribed burning and site preparations for new seedlings.
Data processing and AI
AI platforms ingest satellite and UAV imagery, growth-model outputs and soil maps to forecast timber volume, carbon sequestration potential and disease spread patterns. By simulating various harvest-rotation scenarios, the system optimizes cutting schedules to balance economic returns with ecological sustainability. AI can also assist with pathing decisions for forestry machinery, for example.
Automation and robotics
Robots are increasingly being used for activities such as planting, pruning and harvesting trees, even in challenging terrains. Automated harvesters process stems on the spot into industry-ready logs of a predefined length and diameter that are prepared for automated log loaders. Stock-keeping and inventory is automatically done while harvesting.
Drones
Drones may have a wide range of functions in forestry, from carrying the sensors necessary to gather data for decision making or modelling to dispersing seed pods coated with nutrient pellets and mycorrhizal spores across burnt or clear-cut sites. They may also be employed for sampling in the canopy or even fitted with cutting implements.