askcust develops satellite field monitoring services on open Sentinel-2 data with a 5-day revisit: cloud masking, vegetation indices and the history of every field over the years. For agronomists and managers who need to see problem areas before driving out to a field, not after.
Satellite field monitoring is the regular retrieval of multispectral imagery for a farm's field boundaries, filtering out clouds and calculating vegetation indices to track crop condition without driving the fields. The service keeps each field's history and alerts when an area stands out from the rest. Unlike NDVI monitoring, which analyses one index and its dynamics, satellite monitoring covers image sources, cloud processing and alerts.
The service compares every field with its own earlier images and with neighbouring fields of the same crop and alerts the agronomist when an area falls behind. Field visits start from the places that stood out on the image rather than from the nearest field.
Clouds and their shadows are removed with each image's mask, so the index is calculated only from clear pixels of the field. If a field is covered by clouds, the service shows the last clear image with its date instead of substituting spoiled data.
Images are clipped to the farm's field boundaries from the digital field map, so the agronomist opens a familiar field with its name, crop and history instead of searching a global map by coordinates.
The Sentinel-2 archive makes it possible to restore a field's condition for previous seasons and compare the crop with what grew there before. For rotation and yield analysis the agronomist sees curves over the years, not a single image.
askcust builds the service on the Copernicus and Sentinel Hub APIs; if commercial imagery is needed, the farm buys it directly from the provider.
As soon as a new Sentinel-2 image appears in the Copernicus Data Space Ecosystem or Sentinel Hub, the service fetches it for the farm's field boundaries, with no manual searching or file downloads.
Clouds, their shadows and haze are removed using the scene classification layer of the Level-2A product; only clear pixels go into the field indices, and the cloud share is shown next to every image.
NDVI, NDRE, NDMI and other indices are calculated within each field boundary: an index map, the mean value, the spread and the share of areas below a threshold.
Seasonal index curves for every field, comparison with previous years and with fields of the same crop, and an image archive over the years for crop rotation analysis.
The service alerts the agronomist when a field's index drops sharply between two clear images or when part of a field clearly lags behind the rest; the message contains a map and a link to the field.
Areas that stand out on the image become inspection points with coordinates for the agronomist or scout; inspection results return to the same map.
Free Sentinel-2 imagery is enough for most fields; commercial imagery makes sense where high detail is needed.
| satellite | resolution | revisit | data cost | best for |
|---|---|---|---|---|
| Sentinel-2 (ESA, Copernicus) | 10 m in visible and near-infrared bands, 20 m in red edge | 5 days with two satellites at the equator, 2–3 days at mid-latitudes | free under the Copernicus open data policy | the agronomist's daily work: indices, trends, field comparison |
| Landsat 8 and 9 (NASA, USGS) | 30 m in multispectral bands | 8 days with the two satellites combined | free through USGS | extra images in cloudy periods; the Landsat archive covers history since the 1980s |
| PlanetScope (Planet) | about 3 m | daily imaging of the land surface | commercial licence; the price per area and period is set by the provider | small fields, trial plots, checks after hail or winterkill |
Imagery comes from open Copernicus and NASA/USGS data; askcust writes the processing and interfaces and does not sell imagery.
| criterion | Off-the-shelf monitoring service | Service by askcust |
|---|---|---|
| Start | Sign up, upload the boundaries, and field images are visible the same day. | Development is needed; the first stage is imagery and indices for the farm's fields. |
| Fields and data | Field boundaries and history are kept in the service, apart from the farm's records. | Fields come from the farm's map, and imagery sits next to records and operations. |
| Indices and rules | The set of indices and alert rules the provider built for all its clients. | Indices, thresholds and alert rules are tuned to the farm's crops and processes. |
| Cost | A subscription per hectare; the fee grows together with the land bank. | One-time development and support; Sentinel-2 imagery is free regardless of area. |
| Links to other systems | Export of images and data within what the service plan allows. | Indices and inspection points go to accounting, the mobile app and reports through an API. |
| When to choose | Only imagery and NDVI are needed, with no link to records or own systems. | Satellite data must become part of the farm's system rather than a separate service. |
askcust collects field boundaries, crops and the questions the agronomist wants answered from imagery: crop condition, damage, moisture, variety comparison.
The image archive for past seasons is loaded for the farm's fields; the agronomist compares index curves with what they know about these fields and so validates the settings.
New images load automatically, clouds are removed, indices are calculated per field, and the agronomist receives the first alerts in the current season and marks which of them are useful.
Alerts become inspection points in the mobile app, and indices become layers of the digital field map and data for the manager's weekly reports.
Updates for changes in data providers' APIs, new indices and alert rules, and new fields connected before every season, by agreement.
examples of askcust's work →not agriculture, but the same kind of task: records, CRM, automation
The main source is Sentinel-2 from the Copernicus programme: 10 m in the visible and near-infrared bands, and the data is free. Landsat 8 and 9 give 30 m and are also free; they are added for more frequent observations. Commercial PlanetScope imagery at about 3 m is paid for, and the farm buys it directly if high detail is needed.
The two Sentinel-2 satellites image the same point at least every 5 days, and at Ukraine's latitudes, thanks to overlapping swaths, every 2–3 days. Clear images are fewer: in cloudy periods a field can be covered on several images in a row. That is why the service shows the date and cloud cover of every image, not just the latest map.
Wait for the next clear image or take data from another source. The service removes clouds and shadows with each image's mask and does not calculate the index from spoiled pixels, showing the field's last clear state with its date instead. Sentinel-1 radar imagery sees through clouds, but vegetation indices such as NDVI cannot be calculated from it.
Yes. The Sentinel-2 archive is available from 2015, so index curves over several seasons can be built for every field and the crops in a rotation compared. For deeper history there is the Landsat archive with 30 m imagery from the 1980s. askcust loads the archive for the farm's field boundaries once, at the start.
Not directly: a satellite shows that part of a field lags behind or changes differently from the rest, but the cause is established by an agronomist in the field. That is why the service turns such areas into inspection points with coordinates, and inspection results return to the same map and become part of the field's history.
The cost depends on the number of indices, alert rules, archive depth and integrations with the field map, records and the mobile app, not on hectares: Sentinel-2 imagery itself is free. Commercial processing API fees, if needed, are included in the estimate. askcust estimates the first stage (imagery, cloud masking and indices for the fields) with a fixed scope.
A reply within one business day; a call only if we agree on one.