June 8, 2017
Written by Ognen Plavevski
This article originally appeared on ICTworks. Reposted with permission.
In February CRS, in collaboration with NetHope, flew a drone, or Unmanned Aerial Vehicle (UAV), over cashew farms in central Benin. Just from a first look at the images we could see immediately that there were several issues the farmers needed to address.
We could see where there is space available to plant more trees and how many could be planted, where trees needed to be thinned out, where there had been burning, and areas for follow-up due to invasive weeds or other problems.
Since then, we have found software that automatically tells us: tree counts, crop stress, and gradient changes that can show where water goes. These are hard to detect with the human eye, yet with this information, the field agent can use his limited and expensive time to pinpoint areas that need an in-person visit.
Over time UAV’s have become indispensable in a lot of industries, and their growth is expected to grow significantly over time. Their smaller size, quick deployment and dropping cost make them ideal for a growing list of areas.
We’ve been tracking UAV use in CRS for some time now, and we have two projects where we used them for emergencies, but this is the first case where a project requested their use for agriculture. We documented our experience on AgriLinks and wanted to share more technical data here.
If you’re thinking of using UAV’s for Agriculture, here’s 11 great reasons why you should consider them for your next project:
1. Lower analysis costs.
You can scan large plots of land in one flight (up to 50 hectares for quadcopters, up to 3,000 hectares for fixed wing) and get imagery resolution that is much higher, and can provide much more details than satellite imagery.At the same time, the cost of UAV’s has dropped significantly. You can buy a starter UAV for around $850.Consequently, there’s a very high return of investment (RoI) for using UAV’s for agriculture. In developed country contexts, the RoI is about 146% using this UAV ROI calculator.
2. Detect crop stress, weeds, and more.
UAVs are getting equipped with a lot more sensors than just a regular camera. You can use multispectral, infra red, etc., which can provide you with a lot more analytics then just the regular camera.
3. Count number of crops.
As we said before, UAV’s can be used to count nut trees, but they can also be used to estimate the number and maturity of almost any crop. We can then use those numbers to predict yields.
4. Get accurate maps of individual fields / segments.
Our project needed to map the boundaries of farms, which were really small. We started doing that manually using ESRI’s ArcGIS Collector App to draw the boundaries while walking the terrain. To map all of them out would have been a monumental effort.
Using UAV imagery we printed the farms on paper and provided them to farmers. We had them draw on top of the farms, and the results were amazingly accurate compared to the walking method. All of that in the fraction of the time it took to walk the farms.
While this is already being done with Google Earth by some projects at CRS, it is often difficult to do this well due to low image resolution. UAV imagery makes it easier to detect farmers’ fields.
5. Track advanced farming practices.
Depending on the farming practice, some can be tracked visually, like in this drone-filmed farming video, while others can be tracked through advanced algorithms such as NDVI. This will tell you if farmers are actually practicing advanced farming.
6. Track change over time.
Since you can take as many flights as you want, you can do a before / after harvest, or before / after storm events. This can tell you if there’s been progress made, or how disastrous a storm was to the farm.
7. Make better-informed decisions for precision agriculture.
Using the information at hand, especially from the NDVI sensors, you can see what areas of the farm are in more need of fertilizer, need less fertilizer, need more or less water. Having this information can change the crop outcome significantly.
8. Analyze elevation, water flow and erosion.
Imagery can tell you how the land is sloped allowing you to modify your farm to optimize water use or minimize erosion.
9. Walk the farm in 3D to see crop height.
Since multiple images of a farm are taken, a 3D view can be generated from the imagery. If we combined this with Google Cardboard, we can have a virtual reality view, and literally walk on the farm as if we were there.
10. Applying pesticides.
UAV’s have been used to apply pesticide and fertilizer for some time now. While the prices of these UAV’s is still high, this is something to watch. As price declines this could be a game changer making precision agriculture a reality in new places in developing countries. This greatly reduces the water needed, reduce waste, 97% reduction in cost compared to applying it to the entire crop, very efficient, and much more.
Certain UAV’s can be used to plant seeds, with uptake rate of 75% and a cost decrease of 85% in developed countries. While the cost comparison would inevitably be a bit different in many developing countries, this can help address areas with labor shortages and high value crops that need precision planting.
This is something to watch and begin to experiment with as prices continue to decrease for UAVs. Economies of scale such as; large development projects, multiple cooperatives, and high quality buyers working with small farmers could make this reality.
By Ognen Plavevski, Architect of ICT4D Solutions at CRS.