Agricultural applications for UAS data

We can talk about how amazing this new technology is all we want but what most people are beginning to ask now is ‘how does it make me more money?’ I thought I would put together a list of some of the more common applications for data collected from small fixed wing UAS, particularly in broad acre agriculture.

1. Scouting

Probably the most talked about, and easiest to apply is the ability to aid in scouting paddocks. Think about the perspective you get when you fly over farm land. The process is something like fly over paddock with UAS using a NIR & visual capable sensor, stitch images together to form a georeferenced mosaic, and then calculate Normailsed Difference Vegetation Index (NDVI) or Difference Vegetation Index (DVI). The resulting map would show the variability of crop health over the entire paddock and allow you to concentrate on areas of poor health. In addition, human driven variation is very obvious such as planter problems, compaction, chemical application etc.

To make this process even easier there are a couple of iPad and Android tablet applications (e.g. PDF Maps) that allow you to import the map and use your GPS position to locate areas on the map you are looking for. You are then able to add notes and photos by putting down place marks on the maps. Below is a great demonstration of PDF Maps by Crop Tech Consulting.

2. Site specific weed control

In a similar process to Scouting, making sure you use a sensor with high spatial resolution, the resulting map could be used to identify where large individual weeds are located in a paddock. This could work in a fallow or in crop situation. In our farming environment we face glyphosate resistant weeds that require high chemical coverage to kill. Again an app like PDF Maps could be used to find single weeds in a paddock to mechanically or chemically eradicate.

Feathertop Rhodes Grass in a tidy fallow
Feathertop Rhodes Grass in a tidy fallow

3. Variable rate applications

Precisely placing inputs where they are most needed rather than blanket applications should increase yield and reduce wastage. Variable rate spreaders, sprayers and air seeders have been around for while now but the uptake use has been less than many expected. A rapidly developed georeferenced NDVI or DVI map from a UAS in combination with in paddock examination of what is causing variability puts you in a good position to generate a suitable VR fertiliser application map. Agribotix discuss their method for producing VRA maps here.

Variable rate application map
Variable rate application map

4. Insurance/Drift/Environmental regulations

I spent time at a few UAS conferences while traveling through the US & Canada and there seemed to be a consistent presence from insurance companies. I can understand why, especially in North America where insurance is such a major part of farming. If an adjustor from an insurance company is able to rapidly access a map that correlates closely with what they see on the ground then they are able to adjust the area much more accurately than a ground assessment alone, which is better for everyone involved.

Accurately mapping areas of drift damage and ability to map areas of environmental concern has similar benefits.

5. 3D/DEM/DSM

Employing the structure from motion technology, digital surface models can be generated from still images collected with a UAS. Trimble discuss the surveying potential of this technology at length in a white paper available here. In summary, with ground control points, survey grade information can be rapidly generated in their photogrammetry software with data collected from a UAS. Agisoft Photoscan and Pix4D offer similar functions.

Bezmiechow airfield 3D Digital Surface Model data from Pteryx UAV
Bezmiechow airfield 3D Digital Surface Model data from Pteryx UAV

6. Plant stand

Corn being a pillar crop in the mid-west USA, many are talking about the ability for UAS data to determine site specific plant stands in row crops. The application being to make the decision to replant or not and to evaluate planter performance.

Future applications to get excited about

Above are applications for UAS data than you can reliably apply right now for a reasonable amount of money. Some future applications include:

  • Thermal
  • Lidar
  • 7+ band multispectral
  • Hyperspectral

Topography maps for drainage, Part 1

Recently I’ve been looking at elevation data collected from our Trimble FMX GPS system we use in our tractor.

I imported the logs from this years planting operation. We plant on a swath width of 12m so the data resolution is quite good. The data logs are then transformed into a digital elevation model or DEM and then interpolated to create smoother surface. From the DEM, so far I have developed a contour map and an exaggerated 3 dimensional surface.

Exaggerated 3D surface – note that it is twisted 90 degrees to the below contour map. You are looking from the western side of the paddock.
Contours 20cm

You may have noticed from these diagrams that this paddock has quite interesting topography. A couple of places in this paddock water fails to drain freely which can sometimes lead to water logging. My next step is to identify these sinks.

Software used is the freely available QGIS with GRASS & SAGA extensions available at http://qgis.org.