Technology

Detecting clouds, shadows and snow

We analyze the satellite photo of the field to identify artifacts on it so that we can use images without them in future processes. This stage is necessary for the correct work of the entire platform because only clean pictures allow you to properly estimate field characteristics.

Dense clouds
Translucent clouds
Cloud shadows

We recognize thick and translucent clouds, shadows and snow

Allocating field boundaries

We manually marked tens of thousands of fields, and then trained an algorithm to allocate boundaries automatically. We show what happens with fields at any scale, from a whole region to a particular piece. As a result, any farmer can receive information about the state of his fields in our platform.



IoU 0.85
The accuracy of the automatic markup model
SVG Created with Sketch.
49 years  — the time one person would spend to manually mark these fields
2018
2019
2020
Our algorithms allocate field boundaries with a 5−meter accuracy
21,603,849
Fields marked in the USA
35,923,503
Fields marked in Europe
Our algorithms allocate field boundaries with a 5-meter accuracy
2018
2019
2020
21,603,849
Fields marked in the USA
Our algorithms allocate field boundaries with a 5-meter accuracy
35,923,503
Fields marked in Europe
49 years  — the time one person would spend to manually mark these fields

Detecting more than 20 crops in the middle of season

With multispectral images, we automatically determine a crop that grows on a field. To clarify the information, we use data from the Sentinel−1 radar satellite.

We have detected crops across the fields of Europe and the United States. As a result, it is easy for farmers to begin using our system — we will automatically show what crops they have. In 2020, we will learn what farmers are growing all over the world.

376,835,301 Ha

The size of ​​the analyzed fields in Europe and the USA

Accuracy of recognition F1score
0.92
0.96
Mid season
End of season
Germany4 352 263 fields with total area of 18 200 467 ha
Wheat
24.3%714 729 fields4 415 611 Ha
Corn
9.8%392 965 fields1 780 392 Ha
Barley
2.5%80 840 fields451 446 Ha
Beet
17.9%762 697 fields3 250 187 Ha
Other
45.6%2 401 032 fields8 302 831 Ha
Germany4 352 263 fields with total area of 18 200 467 ha
Wheat
24.3%714 729 fields4 415 611 Ha
Corn
9.8%392 965 fields1 780 392 Ha
Barley
2.5%80 840 fields451 446 Ha
Beet
17.9%762 697 fields3 250 187 Ha
Other
45.6%2 401 032 fields8 302 831 Ha

Determining the sowing date and plant phenostage

With data from the Sentinel−1 satellite, we find the sowing date. Then we determine the stages of plant development using multispectral images. It helps the farmer to choose the best time for application of fertilizers and pesticides.

5 days

The accuracy of determining the plant phenostage

Aiming to automate agriculture in its entirety

OneSoil platform users add information about their fields to the system. Thanks to this, we are constantly improving our algorithms. The more data there is, the more accurate our recommendations are. And consequently, the farmer’s job becomes faster and more efficient.

We believe that agriculture will be fully automated. Special programs and applications will collect information about the field, process it and make decisions with minimal human involvement. We are already building a system of this kind.

Soon we will be able to:

  • Predict
    the yield
  • Plan and monitor fieldwork
  • Give recommendations to the farmer at all stages of work
  • Predict plant diseases and the emergence of pests

Email your questions and suggestions

hello@onesoil.ai