Agricultural UAV's and crop analysis
September 8, 2014 1:42 PM   Subscribe

I'm interested in learning more about precision farming and the use of drones / UAV's in crop analysis. I've done plenty of online research but I'm hoping to benefit from the experience of the hive mind.

My family has a banana / plantain farm and, from what I've researched, this technology has amazing potential.

Has the technology been used or tested in tropical farmlands and mountainous terrain for crops such as coffee and tropical fruits? What are some good online resources for someone who would like to learn more and eventually use this technology in their own plantations? Do you have any first hand experience using UAV's in agriculture that you could share? How affordable is the technology? How is the data gathered and and how steep is the learning curve to learn to interpret it?

This would be for use in South America.

Thanks!
posted by juva to Technology (3 answers total) 5 users marked this as a favorite
 
Here's an interesting website about UAV use for agriculture if you haven't seen it.

I haven't used them personally, so this is from what has been passed on to me: my family started using drones on their hay, wheat, feed corn, and alfalfa fields last year, I believe working with a local community college due to the FAA restrictions. They use it primarily for spotting problems (weeds/pests/disease) and for irrigation (i.e. identifying areas where water is pooling or isn't reaching). My dad has commented that the photos are good enough to spot and identify individual weeds. They've programmed the drones to fly the same paths, so it's really useful for spotting daily, weekly, and monthly problems. My dad has gotten really interested in getting some infrared technology next year but for what specifically I don't know.

I know of two main problems they've had in actually using them:

a) Learning to fly the drones themselves. This was definitely not as easy as it appears - you have to have a basic understanding of the mechanics of flight and radio control models. All of the operators spent lots of time in flight simulators to get better at using the drones, and yet going on their second year they're still having problems with getting good photos due to flight problems (i.e. the wind causes the UAV to roll) and crash landing them. The area where they live is quite windy and has a number of weird thermal currents due to being in the foothills of the mountains, which would be of interest to you. My dad has commented that they've learned to fly them only at certain times of the day as katabatic and anabatic winds are a problem.

These crash landings have caused them significant time and money. (A further complication was figuring out a "runway" which was flat enough to take off and land.) Another challenge is that it really requires at least 2 people, it's boring as hell when done properly, and requires line of sight at all times for safety reasons, which can be a problem in hilly terrain. This was enough of a problem that they stopped using the drone to check on their livestock.

b) Data acquisition, use, and storage. When they do daily flights to take photos they have to stitch the photos together in Photoshop, which often results in a photo way too large to use properly, and then they have to figure out how to use it in a meaningful way. It can be quite overwhelming on a day to day basis. IMHO my folks haven't quite figured this out yet and suffer from data overload. In addition, you have to have a good understanding of how GIS programs work, including having the GIS software - I recommend ArcGIS from Esri. That alone can be a steep learning curve.

That said, my dad is positive that getting to problems, particularly insect and water ones, significantly increased their yield for two years now.

My parents haven't said much about the programs and drones they are using, but I do know it was quite an investment, between 1,000 - 5,000 dollars, and that may only have been on their end. I think they are using CropCam. I also know they love it and it's been very useful to them.

You don't have the FAA, but there may be similar restrictions in your area.
posted by barchan at 2:42 PM on September 8, 2014 [2 favorites]


My work overlaps somewhat with some precision ag people, and we use drones sometimes in my work, but I'm not an expert in either precision ag or drones.

The precision ag people I know are not using drones, and instead are using equipment-mounted gps and data collection equipment, for a few reasons that may guide where you look for information on this: First, farms here are large, and only the larger farmers can make the upfront investment in precision ag, but as a general rule the size of areas that make sense for smaller quadcopter drones (the only kind I've seen used in the field) is smaller. Second, they are already running equipment over every inch of their fields already, so it's relatively straightforward to add sensors and so on to a farmer's existing equipment, rather than have them acquire and learn to use a drone. Third, in this area at least, there is free and reasonably high resolution aerial imagery available from the government, updated every year or two, so the precision ag needs are more on the data side (soil nutrient levels, fertilizer and water application rates, seeding, etc) than for imagery.

All of those are different from the kind of farming you are describing, where the imagery from a drone might be really valuable. I'd just be careful that if you are looking at best practices, you are drawing on examples from similar size and terrain farming; the large, relatively flat-land farming of much of the US (or parts of Brazil, for that matter) have really different data needs and options.

The drone imagery we use in my work is really valuable for assessing conditions before and after implementing a project, because as barchan notes the quality of the images is good enough to identify individual plants and rocks. It's also repeatable, because the drones fly predetermined routes and take photos at set intervals. But for anything other than just showing people a cool photo, the imagery needs work -- either to be orthorectified for GIS, or for analysis of change over time, or anything else. (The gps-linked data the precision ag guys get drops immediately into mapping and those irrigation-CAD programs, as well as the GIS programs that the FSA and NRCS use for field mapping and conservation planning, which again may not be an issue where you are.)
posted by Dip Flash at 5:06 PM on September 8, 2014


Thank you for the info! It's been really valuable and a great starting point.
posted by juva at 2:59 PM on September 9, 2014


« Older Where to get yesterdays UK papers in NYC?   |   How big is the head of a pin? Newer »
This thread is closed to new comments.