The Agriculture Drone Scene: Challenges, and New Tech to Solve (Some of) Them

There are a number of drone companies out there right now aiming to help farmers become more efficient by using data to optimize efforts and avoid waste.

Raptor Maps CEO Nikhil Vadhavkar frames the project of “ag drones,” as they’re often called in the industry, pretty succinctly in this quote:

“Worldwide, about a third of all crops are lost. It’s easy to blame bugs, but really, a lack of information is destroying these plants.”

As the competition to make the best ag drone stiffens (even DJI got in the game a little while back), the improvements offered by various startups have looked to tackle different aspects of the biggest challenges faced by farmers.

Ag Drones

Image Source

If a lack of information is truly to blame then drones can certainly help. But identifying what information is needed, getting it, then actually using it to make improvements is a lot trickier than you might think.

Below are some of the biggest challenges faced by farmers looking to use ag drones, and possible solutions offered by new technology.

What Are Some of the Challenges Ag Drones Face?

1. Help, I’m Drowning in Data!

One of the primary challenges for farmers, as for frankly all of us these days, is about data—or information, as Nikhil Vadhavkar refers to it above. Data is great, but figuring out exactly what data you need and filtering out all the extraneous data can be really challenging, especially at a time when we are often being told that more data is a good thing.

The truth is, data for data’s sake is not at all a good thing. We could be drowning in data and still be no further along toward solving the problems we initially set out to solve if we don’t first establish what data we want, and why we want it.

To pick one example among many, a study conducted by North Dakota State evaluating the use of drones in agriculture set out to investigate:

  • stand counts in corn, sunflowers and sugar beets;
  • the effectiveness of nitrogen applied to corn and wheat;
  • iron chlorosis deficiency in soybeans;
  • and yield predictions for corn, soybeans, wheat, sugar beets and sunflowers.

These data points seem fairly simple, until you dive a little deeper and realize that the images collected by the study’s ag drones will need to be carefully reviewed in order to draw conclusions about items like ‘nitrogen effectiveness’ and ‘iron chlorosis deficiency’.

For starters, you’ll need a specific definition of each item on the list, and and criteria for the data required to establish things like ‘effectiveness’ and ‘deficiency’. This way you know what you’re looking for—which data is important, and which can be ignored.

Once you establish all of these criteria and actually collect the data, you’ll need a review process to sift through the raw data, and turn it into useable data. That review could take hundreds of person-hours, or it could be automated. But how do you automate? Is it possible to do so in every instance?

We won’t dive any deeper, but you can see how data is a challenging area for agricultural drone efforts, both in terms of identifying what and how much data to collect, and also in terms of analyzing it after you’ve got what you want.

2. Precision and Cost

Let’s say you do get the data piece taken care of, and hone in on the crucial data points for your operation as well as how to collect and assess that data in a timely manner.

The next challenge is that the data collected is only as good as the tools being used to collect it.

So you need to make sure your cameras have excellent resolution, and can pick up changes in color and texture from a distance (since those factors will be two of your biggest indicators of soil and plant health). Also important is that your drone handles well, and can actually help you collect the information you need.

And finally, are you doing this in-house, or hiring a pilot? And at what point will the data collected be too expensive to justify—that is, when will using ag drones cost more than the money you anticipate saving by avoiding waste in your crops?

3. Implementation, or Getting ‘Er Done

So now let’s assume you’ve taken care of the two challenges above, and you got some great, precise, actionable data. Now it’s time for you, as a farmer, to take action and save those crops.

Just think about it—one third. It should be easy to allocate time to save one third of your crops!

But this too might be more complicated than it seems on the surface.

Farming occurs in seasons, so even though you might have excellent ideas to implement, you may have to wait some time before you can get up and running.

And even if you do learn something you can hypothetically implement right now, it’s still only as actionable as the crew you have and the labor hours you can put toward that improvement (and, when you consider the hours required simply for routine maintenance tasks, adding new tasks may prove challenging).

Also, how do you really know whether all this effort will save you a third? What if it only saves you 10% of your crop, and all the time and money you put into the effort just helped you to break even?

Some New Solutions

None of this is to say that agriculture can’t be improved by drones, or that we shouldn’t be trying.

Our goal here is simply to call out the inherent difficulties faced both by farmers trying to use ag drones, and by ag drone startups themselves.

And in fact there is plenty being done to tackle the challenges inherent in running a successful ag drone operation. Below are some solutions to these challenges, presented by a few select ag drone startups.

1. Data Solutions

When it comes to the first challenge regarding data, Raptor Maps has created a platform that helps farmers hone in on the data points that matter, and find actionable data.

Here is a quote from a recent post we wrote about them:

Raptor Maps goes beyond just drone mapping and yield data overlays. Since farmers are constantly experimenting with new chemicals, fertilizer regimens, seed types, and techniques, Raptor Maps allows them to perform statistical analyses for A/B testing to compare and optimize these various inputs for fruits and vegetables.

That’s pretty neat, and sounds pretty useable to us.

Other companies are doing similar things, all toward the goal of finding the right data, and presenting it in a user-friendly, actionable manner so that farmers don’t end up drowning in a sea of information.

2. Precision and Cost Solutions

Ag drone startup Gamaya is using new hyperspectral imaging technology to help farmers get better and more accurate data.

According to Gamaya, hyperspectral imaging is 10x more powerful than the standard multispectral technique used by the vast majority of drones in the ag drone startup world.

Gamaya Ag Startup

Image Source

Here is a quote from a recent press release from Gamaya about the hyperspectral imaging approach, and why it’s a game changer for ag drones:

Many Ag drone startups use off-the-shelf technologies, while Gamaya has developed its own state-of-the-art cameras, as well as the technology necessary to interpret the images. Hyperspectral imaging is much more difficult to implement, and is the only imaging format that enables the use of artificial intelligence for crop yield forecasting, and accurate application of pesticides and fertilizer, and other inputs such as labor and water.

This infographic from Gamaya is also useful for understanding the power of hyperspectral imaging, and what it means for farmers:

Gamaya ag drone

3. Getting ‘Er Done

Unfortunately drones may not be equipped to help with this part for a long time. For this, your best bet is to eat another plate of collard greens, put on your boots, and get to work! 🙂

On a more serious note, we recognize that the way we’ve presented challenges and possible solutions in this article is somewhat over simplified. All of the challenges faced by ag drones are interconnected, and we most likely won’t find a silver bullet that addresses every single one of them.

Watching drone startups tackle the problem of big data, and how to collect data and sort it into actionable advice, presents one more fascinating iteration of the constant dilemma we’re faced with in this new golden age of technology.

Which is: How do we create new things that aren’t just neat, but will actually improve the way we live our lives?

Alan Perlman

Founder at UAV Coach
Alan is an FAA-certified drone pilot and founded UAV Coach in 2014 to help connect drone enthusiasts, to provide world-class sUAS industry training courses, and to help push the drone community forward with a focus on safety and commercial opportunities.