DroneCon recap vid: Daniel McKinnon, “What Do Farmers Really Want?”

Here’s the latest of our DroneCon 2014 recap vids: Daniel McKinnon on UAVs in agriculture.

Here’s his abstract:
The technology community has embraced UAVs as the solution to all of agriculture’s problems and inefficiencies. Drone-collected images can theoretically lead to reduced fertilizer, water, and chemical application; increased yields; reduced uncertainty; and ultimately increased profits for the farmers. However, while many of these hypotheses may be true, the advanced data currently being collected by drones is far ahead of the adoption of the precision agriculture technologies that enable farmers to act on the data. Through working with farmers and agronomists over the last 18 months, Agribotix has identified the types of data farmers are interested in and a number of different applications for UAV-collected imagery that any farmer can use immediately without strongly deviating from a standard workflow.

Views: 1511

Comment by Mike T on July 2, 2014 at 6:09pm

video is private

3D Robotics
Comment by Roger Sollenberger on July 2, 2014 at 6:10pm


Comment by Mike T on July 2, 2014 at 6:31pm

thanks for uploading :)

Comment by Kevin Hester on July 2, 2014 at 8:32pm

That was my favorite talk of the convention!  Thanks for giving it.

Comment by Gary Mortimer on July 3, 2014 at 1:18am

Realistic I like! This is a fantastic talk. I was speaking to a guy last night that who is taking his push pull Cessna several thousands of km to survey a huge area with a Lidar and I was floored when he said he was only being paid $1 a hectare for the processed data, aircraft time and accommodation. His over heads must be huge, if he is prepared to go out with a manned aircraft and commercial lidar at that rate how could I compete with simple data... 

Comment by Gerard Toonstra on July 3, 2014 at 4:31am

Very good presentation, state-of-the-art in what you do with collected data using affordable systems.

What you can do with multirotor is increasing. 35 minutes is well within reach now and also affordable. You'd want to work with those if you have complicated areas (lots of trees) or just a small farm where you don't want to fly over adjoined patches. its slower speed allows you to make much more detailed ortho's than fixed wing because you can fly a lot lower, so there are definitely uses for it.

You can chop up large orthos using some tools: http://www.gdal.org/gdal2tiles.html  and tilemill for example. This allows you to manage the amount of data sent across or otherwise how many images are kept in memory at the same time.

Comment by John Stuart on July 3, 2014 at 6:43am

+1 for the improving prospects for multi-rotors, as the technology allows them to push flight times. In vineyards too, its not easy to find a decent spot to land a fixed-wing.

Regarding the economics of the service, I found it interesting that you started out with offering the services for free, I've had to do the same thing and hope that the data is useful enough for the client to buy in at some point.

Thanks for the presentation, its great to hear what you guys have already achieved in little over a year. Your willingness to share your experiences is truly appreciated.

3D Robotics
Comment by Daniel McKinnon on July 3, 2014 at 10:04am

Thanks for all the great feedback! DroneCon was a lot of fun and I was really glad to have been given the opportunity to present.

There have been a bunch of questions about economics and the answer is a little bit tautological. The data is worth the value it creates for the farmer. The small minority of farmers who buy satellite imagery pay around $2.50 an acre from Digital Globe, but the data are generally too dated for high value applications like fertilizer prescription maps and the farmers are left to do the interpretation. Crop consultants typically charge $10-$15/acre to provide general field recommendations that are based on weekly walk-throughs, without the spatial precision given by UAV-collected imagery. To really extract revenue out of imagery in agriculture, it's critical to show value to the farmer. We've found that value in, to a lesser extent, identifying weeds and quantifying coverage and, to a much greater extent, generating prescription maps for precision applicators.

The software packages currently available to stitch images require a significant sidelap on canopied fields, which typically creates a linear flight of around 25 km for a quarter section (160 acres). We would love to use a multi-rotor for the ease of takeoff, landing, and planning, but have not encountered a system with this kind of endurance yet. I think future solutions for image processing will eliminate the need for so much sidelap and could allow us to reintroduce a multi-rotor, but that will be a little way down the road.



Comment by Andrew Rabbitt on July 4, 2014 at 8:31pm

Fantastic presentation Daniel!  Congratulations.

So, fixed-wing quasi-VTOL with 25+km range is a good target for a useful airframe/system?  That should be doable.  I assume you're carrying 2 x consumer cameras with this - a 350g-ish payload?

3D Robotics
Comment by Daniel McKinnon on July 7, 2014 at 6:53pm

Hi Andrew,

Sorry for the delayed reply, but your idea would be perfect for us. We are fairly airframe agnostic and have spent most of our time figuring out how to process the images and apply them to solve agricultural problems, but what you described would be the hot setup.

We have found essentially identical NDVI results using the red channel, the green channel, or the blue channel, so we only fly a single camera now, making the payload significantly lighter.




You need to be a member of DIY Drones to add comments!

Join DIY Drones


Season Two of the Trust Time Trial (T3) Contest 
A list of all T3 contests is here. The current round, the Vertical Horizontal one, is here

© 2018   Created by Chris Anderson.   Powered by

Badges  |  Report an Issue  |  Terms of Service