FarmPortal - intelligent decision support with AI

AI in agriculture.Field data. Better decisions.

8 AI features in FarmPortal

Where should you target fertiliser? When should you plan spraying? Which field needs checking first? FarmPortal combines IoT sensor readings, satellite imagery and farm history with agronomic models and the AI algorithms we are developing to support everyday farm decisions.

FROM INFORMATION TO ACTION

The context of each field

  • Satellites
  • Soil
  • Weather
  • Field history
  • Machinery
  • Costs
  • IoT sensors

What is AI in agriculture?

Artificial intelligence in agriculture uses models that learn relationships in data to support crop assessment, forecasting and decisions. At FarmPortal, it complements a decision support system (DSS) that also uses agronomic calculations, weather models and device data.

Practical support for everyday work

Available analytical tools and DSS help turn data into action. The scope of analysis depends on the connected data sources, devices and services.

Agronomic assistant

Which field should you check first?

The DSS brings together crop, weather and treatment information. Alerts and field history help you choose where to focus your attention.

In practice

Before heading out, you check wheat alerts and plan a field inspection.

Explore this feature: Agronomic assistant
Fertilisation and NPK

Where should you adjust the rate?

Nutrient calculations and satellite zoning connect crop needs with soil test results and crop condition. They help prepare rates for different parts of a field. Variable-rate prescription maps can be used with a compatible ISOBUS terminal and implement supporting TC-GEO to adjust application rates automatically according to GPS position.

In practice

You prepare fertilisation zones, and a compatible ISOBUS spreader adjusts the rate according to the prescription map and GPS position.

Explore this feature: Fertilisation and NPK
Soil sampling

How should you plan soil sampling points?

Automatic field division and point planning organise soil sampling. Previous tests and maps help assess variability within the field.

In practice

Before testing, you define sectors and save sample locations.

Explore this feature: Soil sampling
Irrigation

Does the crop need water?

The water balance, crop evapotranspiration (ETc) and weather data support assessment of irrigation needs and water deficit risk.

In practice

Ahead of hot weather, you compare the water balance with forecast rainfall.

Explore this feature: Irrigation
Spray timing

Are conditions suitable for spraying?

Delta T, temperature, humidity, wind and the rainfall forecast help assess spraying conditions in the context of planned crop protection.

In practice

You check wind and Delta T before sending out the sprayer.

Explore this feature: Spray timing
Disease risk

When should you inspect the crop?

Disease models, available through services including METOS, assess conditions favourable to infection. They help plan scouting and assess the need for crop protection.

In practice

After infection risk rises, an orchard grower inspects the relevant block.

Explore this feature: Disease risk
Crop condition

Which part of the field needs scouting?

NDVI, NDRE, MSAVI, RECI and NDMI indices, together with imagery history, show changes in crop condition. A difference on the map identifies an area to inspect.

In practice

A falling index in one field corner directs you to a specific location.

Explore this feature: Crop condition
Fuel and machinery

What explains unusual fuel use?

Telemetry reports and recorded fuel anomalies help investigate consumption, stops and fuel level changes, depending on the device capabilities.

In practice

You compare tractor consumption and investigate a prolonged idle period.

Explore this feature: Fuel and machinery

IoT sensors — measurements where you work

The Internet of Things (IoT) connects measuring devices to the farm system. FarmPortal uses data from compatible sensors to complement satellite imagery and forecasts with conditions in the field, on machinery or in storage.

Weather stations

Selected METOS, Atmesys and Weenat stations provide air temperature, humidity and rainfall, plus wind and solar radiation depending on the equipment.

In practice

Before spraying, you check conditions measured near the crop.

Soil probes

Volumetric water content (VWC), temperature, water potential or electrical conductivity (EC), depending on the probe, reveal conditions in the root zone.

In practice

You check whether irrigation water has reached root depth.

Frost sensors

Air temperature and wet-bulb and dry-bulb readings from selected devices support frost risk monitoring and alarm thresholds.

In practice

At night, you monitor falling temperatures in an orchard block.

Leaf wetness sensors

Leaf wetness and temperature readings in supported configurations supplement the inputs needed by selected infection risk models.

In practice

After prolonged leaf wetness, you check disease risk and plan scouting.

Storage sensors

Compatible temperature, humidity and CO₂ sensors help monitor storage conditions and changes in the microclimate.

In practice

A temperature rise draws attention to a specific batch of stored produce.

GPS and ISOBUS telemetry

Location, working time, engine speed, load and available fuel readings connect machine operations with equipment utilisation analysis.

In practice

You compare fuel consumption with the tractor’s route and working time.

Explore sensors and compatible devices

How does FarmPortal connect sensor, satellite and farm data?

Device and service integrations, including APIs, bring local measurements together with crop history, treatments and costs. This provides context for balances, alerts and decision support models.

  1. Connect a device or service

    Use a supported integration for a weather station, probe or telemetry system. For a custom implementation, we assess the device model and available API.

  2. Combine readings with field history

    Bring readings together with crop information, satellite imagery and completed treatments. Compare changes over time and conditions at each location.

  3. Use the result in your decision

    Review the water balance, spraying conditions or infection risk. Combined data supports DSS and the development of proprietary AI models in ProfitCast.

Explore data integration options

PROFITCAST · IN DEVELOPMENT

Proprietary AI algorithms. From yield to profitability.

At Agri Solutions, we are developing proprietary algorithms in the ProfitCast project. We combine Earth observation, agronomic and economic data to forecast outcomes at field level. The three solutions below are part of ongoing research and development.

Yield Intelligence

Yield forecasts throughout the season

The model is being developed to update yield forecasts as new satellite imagery, weather and crop data become available.

Before harvest: estimate output to plan transport and storage.

Fertilization Strategy

A fertilisation strategy matched to your budget

We are developing model-based NPK estimates and fertilisation scenario comparisons using Earth observation and reference soil tests.

On a limited budget: compare which field may justify additional fertiliser.

ENP Calculator

Expected net profit for each field

The Expected Net Profit calculator is being developed to combine forecast yield, selling price and costs to compare production scenarios.

Before another treatment: assess how the extra cost could affect the forecast result.

Explore ProfitCast and its development stage

How does data inform a decision?

  1. Connect your data

    Field history, soil tests, satellites, weather, sensors and costs provide the context for analysis.

  2. Review the analysis

    Maps, balances and models show variability, risk and possible actions for a specific field.

  3. Make your decision

    Compare the result with scouting, local conditions and production goals. Record the action you take.

Models grounded in farm conditions

We develop algorithms around specific decisions: rate, timing, location and cost. Our AI development works alongside existing agronomic tools and integrations, with methods matched to each type of analysis.

Data from your farm
The same weather can mean different risks for different crops and treatment histories.
Results to verify in the field
Satellite data shows a change in condition. Scouting helps identify its cause.
Agronomy connected with economics
Within ProfitCast, we are developing yield assessment alongside input costs and expected profitability.

AI in agriculture — questions and answers

FarmPortal offers analytics and decision support for fertilisation, irrigation, crop protection and machinery operations. In ProfitCast, we are developing proprietary AI models for yield forecasting, fertilisation strategies and field profitability. ProfitCast is currently in research and development.

Analysing soil tests, crop condition and field history helps match fertilisation to the needs of individual zones. FarmPortal provides nutrient calculations and satellite zoning, while ProfitCast is developing comparisons of yield and cost scenarios. Results depend on farm conditions and data quality.

Disease models assess conditions favourable to infection and the risk of specific diseases. FarmPortal supports them through services including METOS. Results help plan scouting; they are not a confirmed diagnosis. Not every predictive model uses machine learning.

No. Satellite imagery and indices show crop condition and variability. The model-based NPK assessment being developed in ProfitCast uses reference tests and does not replace laboratory soil analysis.

ProfitCast brings together yield forecasting, fertilisation strategies and the Expected Net Profit calculator. The aim is to compare expected revenue with costs at field level and assess alternative scenarios before harvest.

FarmPortal supports decisions, while the farmer or adviser assesses the result alongside observations and field conditions. Crop protection decisions must take the product label into account. Analysis quality depends on the available data, its timeliness and the scope of the particular model.

FarmPortal integrates selected weather stations, soil probes, frost sensors, storage microclimate sensors and GPS/CAN telemetry. Supported configurations may include leaf wetness measurements. Available parameters and data transmission frequency depend on the device model and integration.

IoT measurements provide information about local temperature, humidity, rainfall or machine operations. FarmPortal can combine them with satellite imagery, field history and treatments for supported DSS analyses. Data integrations also support the development of AI models in ProfitCast. Data exchange with a particular device depends on the available interface, API and configuration.

Start with your own field data

Organise your fields, soil tests and treatment history in FarmPortal. Choose the tools that will support your next decision this season.

AGRI SOLUTIONS SERVICES

Custom AI projects for agriculture and the food sector

We deliver bespoke AI projects for farms, agri-food businesses and research partners. We combine proprietary algorithm development with data acquisition, device integration and results visualisation — from prototype to deployment.

Data acquisition and integration

We connect IoT sensors, machinery and systems through available interfaces and APIs. We organise measurement data, production records and activity history for the project.

Proprietary AI models and algorithms

We design AI/ML models for a specific problem, prepare the data and validate the results. We tailor the work to the business objective and available measurements.

Results visualisation and deployment

We present results in dashboards, maps and reports. We integrate analysis into users’ workflows and the partner’s systems.

FarmPortal and FoodPass — ready-made foundations for your project

In bespoke projects, we use our existing FarmPortal and FoodPass platforms for data collection and visualisation, as well as device and system integration. We extend them with features and algorithms tailored to the partner’s needs.

Experience in EU and NCBR projects
We carry out research and development projects co-funded by the European Union and Poland’s National Centre for Research and Development (NCBR). Our experience includes FoodPass, crop analytics, IoT integration and animal welfare monitoring.
An interdisciplinary team
We combine expertise in agronomy, data science, software engineering and IoT. Together, we translate a farm or production-site problem into data, a model and a tool for the user.