Our ML models have been trained since 2017 with vast amounts of ground truth data.

top-quality.png

Top quality 

We use satellite imagery to ensure our data's global coverage and improve our model every day to provide the best accuracy worldwide. Check out our [demo map](https://map.onesoil.ai/2020/#2.3/44.35/-43.66) for details.

global-coverage.png

Global coverage

We cooperate with agricultural holdings around the world and get confirmed field data from different regions. We are committed to contributing to sustainable agriculture. We run thousands of trials with farmers and holdings to increase yield while reducing input usage.

improvement.png

Continuous improvement

Des algorithmes de machine learning à la pointe du progrès.

Limites de parcelles disponibles pour 59 pays. 
Données sur les cultures pour l’hémisphère nord, l’Amérique du Sud, l’Afrique du Sud, et l’Australie.

Données internationales

Limites de parcelles, détection des cultures, et zones de productivité basées sur l’imagerie satellite et un modèle de machine learning entraîné avec des données terrain réelles collectées depuis 2017.

Données précises

Formats GeoServer API, GeoJSON, GeoTIFF, fichier GeoPackage, ou sous forme de tableau.

Formats de fichiers pratiques

Check out our interactive map here. You can get valuable agricultural insights in sandboxes for different regions of the globe. Sign up and get even more free data for 2020.

Subscribe to get data from the latest season.

Get the best price for your area of interest

We've invested a lot of time and effort into delineating field boundaries on satellite images. We are also fully committed to seeking out ground truth data, training our machine learning algorithms, and producing our own devices. 

We conduct experiments in partnership with agricultural professionals to test out agronomic hypotheses. We've already experimented on more than 50,000 hectares. We also receive speedy feedback about the accuracy of our algorithms from users of OneSoil apps. Our technology is what powers them.

We have field boundaries for 59 countries and crop data for the Northern Hemisphere, South America, South Africa, and Australia.

For 12 crops: Wheat, corn, barley, soybean, rapeseed, sunflower, sugar beet, sugar cane, rice, cotton, sorghum, and grass.

We can conduct all kinds of analyses based on our data. If you want to discuss this, please fill out this [Google Form](https://docs.google.com/forms/d/e/1FAIpQLSdTbvovgLgcFCgGBTKPV8TJ9romOdATTILI_fP0IMwvwEzEiA/viewform). 

Got questions? We have answers!

large_onesoil-map-large.webp

medium_onesoil-map-large.webp

small_onesoil-map-large.webp

thumbnail_onesoil-map-large.webp

onesoil-map-large.webp

Des solutions clé en main pour différents types d’activités. Démonstration gratuite. Tarification ajustable.

Les produits OneSoil reposent sur nos modèles propriétaires de machine learning. Depuis 2017, nous avons développé des ensembles de données établis sur des milliers de points de données terrain en provenance de grandes sociétés agricoles et nous avons amélioré la qualité de nos données.

field-boundaries.webp

Save time and money on manual field delineation. Import automatic field boundaries to your digital ag platform and improve user onboarding. Knowing field borders also increases the accuracy of all further calculations. 
We estimate field borders with a resolution of 2-5 meters even though we use 10-meter resolution Sentinel-2 images. We can do this thanks to our accurate ground truth data and a few scientific tricks.

### Why OneSoil?
**Available globally**

We offer state-of-the-art field boundaries for more than 59 countries, covering most of the arable land worldwide. 

###### **Reliable remote sensing data for each season since 2017**
###### 
We have trained our proprietary machine learning algorithms to improve data accuracy year after year to offer highest possible accuracy. 

###### **5 m accuracy**
######Thanks to years of neural networks training, we provide field boundaries as precise as 5m with minimum field size of 0.5 ha (1.25 acre). 

**Up to 0.96 of IoU (Intersection over Union) score**


Accuracy of the automatic field delineation model is vital for your users’ experience. Try out our field boundaries model now and see how they can contribute to your digital ag platform.

### Data delivery methods:
###### GeoJSON,
###### API,
###### Geoserver API, GeoJSON, GeoTIFF, Geopackage file or tabular format.

Les parcelles détectées automatiquement sont le point de départ de toute statistique agricole et facilitent l’intégration de nouveaux utilisateurs sur les plateformes d’agriculture numérique. 

Résolution 5 × 5 m pour les limites de parcelles
Précision jusqu’à 0,85 IoU

Save time and money on manual field delineation with OneSoil Field Boundaries.

AI-Detected Field Boundaries for 59 Countries

Limites de parcelles

crop-detection.webp

Get a reliable source of crop identification data from all seasons on a global scale utilizing crop rotation info since 2017.

### We work with 12 primary cash crops

Wheat, corn (maize), barley, soybean, rapeseed/canola, sunflower, sugar beet, sugar cane, rice, cotton, sorghum (milo), and grass.

### Why OneSoil?

Using a ground truth dataset that originates from thousands of field trials performed together with several ag holdings, we prepared models that work quickly and efficiently in the Northern and Southern Hemispheres, taking into account all regional features. 

Our proprietary machine learning algorithms have been improving since 2017, with thousands of ground truth data points collected and used for training our model. 

Our crop recognition accuracy has 
**an F1 score of 0.92–0.96**.

### Data delivery formats:
###### GeoJSON,
###### API,
###### GeoServer API, GeoJSON, GeoTIFF, GeoPackage file, or Tabular format.

Crop Identification.png

Identification des cultures partout dans le monde d’après les observations satellite depuis 2017.

13 cultures de rente
Résultat F1-score de 0,96

Accuracy of crop detection starting from 92%. 12 major crops. Improving since 2017. 

AI-Detected Crops Available Worldwide

Détection des cultures

productivity-zones.webp

Improve your FMS system with best-in-class productivity zones that represent field areas of low, medium, and high yield. 

Calculated with satellite imagery for the last six years
Based on NDVI, soil organic matter content, field relief, and yield data. 

![productiivity-maps.png](https://os-strapi-assets-devoted-silkworm.s3.eu-central-1.amazonaws.com/productiivity_maps_56316e19db.png)

### Why OneSoil?
Based on years of analysis of field data and satellite images, we can tell which fields are suitable for VRA technology. This is a prerequisite for digital agronomy and a vital step to ensuring accurate management zones and prescription maps for in-field operations.

### Data delivery formats:
###### GeoJSON,
###### API,
###### GeoServer API, GeoJSON, GeoTIFF, GeoPackage file, or Tabular format.

Productivity Zones.png

Zones de productivité faible, moyenne ou élevée créées automatiquement dans toutes les parcelles, d’après six années d’observations satellite.

Improve your FMS system with best-in-class productivity zones that represent field areas of low, average, and high yield.

OneSoil Productivity Zones

Zones de productivité

field-score.webp

Evaluate field trial results and predict yield in the current season! The field score is a metric for comparing the performance of fields, regions, and seasons against one another by utilizing NDVI and other environmental factors.

### For trading companies
With a field score, you can make yield forecasts compared to a previous season.

### For Seed Companies
You can compare several fields and use the field score as evidence of a successful field trial, comparing different varieties to showcase to your customers. 

### For Fertilizer Companies
You can use the field score to compare yields after field trials. The field score can be used as evidence of your fertilizer's effectiveness. 

### Why OneSoil?
Field scores are based on the Normalized Difference Vegetation Index (NDVI). It helps assess the quantity and quality of vegetation in a field during the vegetation period. 

Lately, we started to combine NDVI and weather indicators to increase the accuracy of the field score.

### Data delivery formats:
###### GeoJSON,
###### API,
###### GeoServer API, GeoJSON, GeoTIFF, GeoPackage file, or Tabular format.

field-score.png

Une évaluation qualitative de la récolte d’après l’indice de végétation normalisé (NDVI), les indicateurs climatiques et la productivité relative de la parcelle.

The field score is a metric for comparing the performance of fields, regions, and seasons against one another by utilizing NDVI and environmental factors.

Field Score To Evaluate Field Trial Results And Predict Yield 

Bientôt disponible : Field Score

cloud-free-ndvi.webp

Get your NDVI data seamlessly, regardless of weather conditions. In some regions, clouds cover the sky up to 50% of the time, which makes conventional NDVI calculations impossible or not accurate enough. With the Cloud-Free NDVI algorithm, we can literally see through the clouds and give you accurate data anytime, anywhere.

![cloud-free-compare.webp](https://os-strapi-assets-devoted-silkworm.s3.eu-central-1.amazonaws.com/cloud_free_compare_5d98ebfc0b.webp)

### Why OneSoil?
**Global reach**
###### We provide NDVI calculations globally, with updates every 3-5 days from Sentinel-2 imagery. Our data is used by customers in Europe, North America, and Latin America. 

**Cutting-edge technology**
###### Our ML algorithms analyze signals from Sentinel-2 and radar data based on Sentinel-1 imagery to improve the visibility of crop vegetation with 95-97% accuracy, even in heavy cloud coverage. 

### Data delivery formats:
###### GeoJSON,
###### API,
###### GeoServer API, GeoJSON, GeoTIFF, GeoPackage file, or Tabular format.

Cloud-Free NDVI.png

Dans certaines régions, la couverture nuageuse est présente 50 % du temps tout au long de l’année.

Les données de NDVI Cloud-free permettent de surveiller l’indice NDVI dans toutes les parcelles, indifféremment des nuages.

With the Cloud-Free NDVI algorithm, we can see through the clouds and give you accurate data anytime, anywhere.

Cloud-Free NDVI To Get Your Data Regardless Of Weather Conditions

Bientôt disponible : NDVI Cloud-Free

sowing-dates.webp

planting-dates.webp

A great tool for validating farmers' reported practices. Our planting dates come into play wherever timing is critical to correctly assessing CO2 emissions. To make accurate calculations, you need to know what was planted and when. Now, it's easier than ever! 

#### Why OneSoil?
By combining signals from two satellites, we were able to make the algorithm operate regardless of the weather conditions present when field operations were performed. 
Based on open data from the Sentinel-1 and Sentinel-2 satellites, OneSoil calculates the dates of field operations with an accuracy of 2 to 5 days as quickly as 3-4 weeks from the planting date.

#### Data delivery formats:
###### GeoJSON
###### API
###### GeoServer API, GeoJSON, GeoTIFF, GeoPackage file, or Tabular format.

Planting Dates.png

Reconnaissance des pratiques agricoles utilisées par les plateformes carbone dans leur processus d’évaluation.

3 à 5 jours correspondent à la précision des calculs effectués 4 semaines après le semis.

Based on open data from the Sentinel-1 and Sentinel-2, OneSoil calculates the dates of field operations as quickly as 3-4 weeks from the planting date.

Planting Dates With An Accuracy Of 2 To 5 Days 

Bientôt disponible : dates de semis

tillage-practices.webp

Our offering for carbon platforms enables satellite-based recognition of CO2-critical operations such as full-till, strip-till, and no-till. 

###### If you would like to monitor the launch of this dataset, please [leave a request](https://docs.google.com/forms/d/e/1FAIpQLSdTbvovgLgcFCgGBTKPV8TJ9romOdATTILI_fP0IMwvwEzEiA/viewform).

![tillage-practices-img.webp](https://os-strapi-assets-devoted-silkworm.s3.eu-central-1.amazonaws.com/tillage_practices_img_7ef89be36c.webp)

#### Why OneSoil?
Using a ground truth dataset that originates from thousands of field trials performed together with several ag holdings, we prepared models that work quickly and efficiently in the Northern and Southern Hemispheres, taking into account all regional features. 
Our proprietary machine learning algorithms have been improving since 2017. 

#### Data delivery formats:
###### GeoJSON
###### API
###### GeoServer API, GeoJSON, GeoTIFF, GeoPackage file, or Tabular format.

Tillage Practices.png

Reconnaissance des pratiques agricoles utilisées par les plateformes carbone dans leur processus d’évaluation.

Our offering for carbon platforms enables satellite-based recognition of CO2-critical operations such as full-till, strip-till, and no-till. 

Tillage Practices For Carbon Platforms

Bientôt disponible : pratiques de labours

cover-crop-detection.webp

Until now, the main source of information on cover crop usage came from farmers' reports and declarations. Get a convenient tool to verify their input at an extremely low cost per field, thanks to satellite imagery analysis. 

![cover-crop--img.webp](https://os-strapi-assets-devoted-silkworm.s3.eu-central-1.amazonaws.com/cover_crop_img_64389defa3.webp)

#### Why OneSoil?
Using a ground truth dataset that originates from thousands of field trials performed together with several ag holdings, we prepared models that work quickly and efficiently in the Northern and Southern Hemispheres, taking into account all regional features. 
Our proprietary machine learning algorithms have been improving since 2017. 

#### Data delivery formats:
###### GeoJSON
###### API
###### GeoServer API, GeoJSON, GeoTIFF, GeoPackage file, or Tabular format.

Cover Crop Detection.png

Get a convenient tool to verify farmers’ input at an extremely low cost per field, thanks to satellite imagery analysis.

Cover Crop Detection For Carbon Platforms

Bientôt disponible : détection du couvert végétale

Produits Global Analytics

OneSoil propose des solutions pour différents types d’entreprises agricoles, adaptées à leurs besoins.

farm-input-companies.webp

Seeking profitable growth in the increasingly digitalized ag world requires drafting the right strategies and making data-driven decisions. OneSoil provides valuable insights for your strategic planning through our product suite dedicated to Farm Input companies. 

Satellite-based analytics results in the best possible cost per hectare and makes it possible to provide a laser-sharp assessment of target market potential. Our suite contributes to your value-add for farmers, helping you lock in regular customers and push out the competition.

#### Integrating field boundaries into apps and internal tools
Optimizing development resources and delivering an optimal user experience via easy onboarding in native apps. 
#### Field boundaries
#### Crop detection
This can be used to plan marketing activities at the regional level, analyze past and current seasons, plan purchases, and build strategies, market opportunity reports, and sales region maps for dealers and territory managers.
#### Productivity zones
Integrating VRA zones into native farm management apps allows your customers to realize the full potential of your farm input products.

#### Data delivery formats:
###### GeoJSON
###### API
###### GeoServer API, GeoJSON, GeoTIFF, GeoPackage file, or Tabular format.

For Farm Input Companies.png

Our suite contributes to your value-add for farmers, helping you lock in regular customers and push out the competition.

For Farm Input Companies | Global Analytics Solutions

For Farm Input Companies: seeds, fertilizers, and chemicals

farm-equipment-oems.webp

Ag machinery producers need to compete both by improving their in-field performance, fuel intake rates, and after-sales care, as well as by shifting to in-field connectivity and precision ag services enabled by onboard digital platforms. OneSoil solutions help farm equipment vendors lock in growers in their eco-systems.
#### Field boundaries
Make it easy for users to onboard and select their fields in your digital ag platform. 5x5m accuracy is not used for precision ag but is a great tool to seamlessly engage users with no need for manually drawing boundaries or uploading them from external files.
#### Productivity zones
Our high-quality productivity zones are based on satellite imagery & calculated using proprietary ML algorithms that estimate zones' productivity relative to each other based on several years of Sentinel-2 satellite images for the field. 

Models were trained and are continuously improved using ground truth data we compiled from more than 1M hectares of fields. 

We compared actual yield figures and NDVI indices in various stages of plant development to identify patterns to build ML models. As a result, our productivity zones show the highest correlation with actual yield data compared to several other leading VRA solutions:

![compare-productivity.png](https://os-strapi-assets-devoted-silkworm.s3.eu-central-1.amazonaws.com/compare_productivity_39df541258.png)
#### Data delivery formats:
###### GeoJSON
###### API
###### GeoServer API, GeoJSON, GeoTIFF, GeoPackage file, or Tabular format.

For Farm Equipment OEMs.png

OneSoil solutions help farm equipment vendors lock in growers in their eco-systems.

For Farm Equipment OEMs | Global Analytics Solutions

For Farm Equipment OEMs

trading-companies.webp

Producers, traders, and users of grains and oilseeds around the world adjust their in-season strategies in line with their yield expectations. 

OneSoil solutions for Trading Companies help predict yield in specific regions and provide crucial insights for decision-making to deal with price fluctuations and production risks. 
#### Crop detection
In-season crop type detection and acreage reporting make it possible to produce yield forecasts based on independent data sources. 

Draw up your strategy for the season and greatly reduce your trading risk with a reliable data partner. 
#### Planting dates
An add-on to our crop detection that enables precise identification of planting dates and provides insights into yield expectations and timing forecasts.
#### Field score
A relative comparison of how particular fields, regions, or seasons perform by evaluating NDVI and climatic indicators. 

#### Data delivery formats:
###### GeoJSON
###### API
###### GeoServer API, GeoJSON, GeoTIFF, GeoPackage file, or Tabular format.

For Trading Companies.png

OneSoil solutions for Trading Companies help predict yield in specific regions and provide crucial insights for decision-making.

For Trading Companies | Global Analytics Solutions

For Trading Companies

carbon-platforms.webp

If you run a platform where farmers can generate carbon credits in return for sustainable agriculture practices, you probably need high-quality data to monitor, report, and verify their declarations. 

Farmers can increase carbon sequestration in soil and reduce greenhouse emissions by applying certain practices such as reducing tillage, adding cover crops, and controlling fertilizer application rates. Carbon platforms need to be equipped with the right tools to generate carbon credits that are in high demand by the corporate world. 

This is where OneSoil solutions for Carbon Platforms add value by increasing the accuracy of their MRV activities.
#### Satellite-based recognition of field boundaries and acreage, crop sequence and seeding dates, tillage practices and cover crops (available soon)
#### Field boundaries
Identifying boundaries and acreage of field(s) where eco-friendly farming methods will be used.
#### Crop detection
Identifying the crop sequence for the last 6 years.
#### Planting and harvest dates
Identifying planting dates and harvest dates automatically.
#### Tillage practices and cover crops
Analyzing the presence of eco-friendly farming methods.


#### Data delivery formats:
###### GeoJSON
###### API
###### GeoServer API, GeoJSON, GeoTIFF, GeoPackage file, or Tabular format.

For Carbon Platforms.png

Satellite-based recognition of field boundaries and acreage, crop sequence, and seeding dates.

For Carbon Platforms | Global Analytics Solutions

For Carbon Platforms

financial-institutions.webp

Crop insurances massively suffer from the hurdles of accumulation risk, moral hazard, and adverse selection. Essential farm data is unconfirmed and can potentially be manipulated, leading to, for example, unrealistic yield data on the actual insured field and insured crop. This results in excessive insurance premiums. 

OneSoil machine learning algorithms combined with satellite data make it possible to project reliable, parameterizable, and scalable data specifically for the insured crop in a specific field. This opens up completely new possibilities. 

Our data can be used by Primary Insurers and Reinsurers for risk and damage assessment: 
◦ Calculating the risk component (underwriting) for both single-peril crop insurance (SPCI) and multi-peril crop insurance (MPCI), 
◦ Increasing the accuracy of forecasts for actual compensations that have to be paid, thus increasing liquidity in case of damage, 
◦ Assessing flood or drought damage, as well as water balance and plant growth models.
#### Field boundaries
Specifying the boundaries and acreage of field(s) where insurance applies.
#### Crop detection
Identifying the crop sequence for the last 6 years.
#### Planting and harvest dates
Identifying planting dates and harvest dates automatically.
#### Field score
A relative comparison of how particular fields, regions, or seasons perform by evaluating NDVI and climatic indicators used for risk assessment.

#### Data delivery formats:
###### GeoJSON
###### API
###### GeoServer API, GeoJSON, GeoTIFF, GeoPackage file, or Tabular format.

For Financial Institutions.png

OneSoil algorithms combined with satellite data make it possible to project reliable, parameterizable, and scalable data for the crop in a specific field.

OneSoil Global Analytics : données agricoles pour toutes les régions

Produits Global Analytics

Limites de parcelles

Détection des cultures

Zones de productivité

Bientôt disponible : Field Score

Bientôt disponible : NDVI Cloud-Free

Bientôt disponible : dates de semis

Bientôt disponible : pratiques de labours

Bientôt disponible : détection du couvert végétale

Des algorithmes de machine learning à la pointe du progrès.

Got questions? We have answers!

Why do you have information about the whole world while other companies don't?

What countries do you have data for?

For which crops?

Can you make forecasts?

Get the best price for your area of interest