June 8, 2016
Yuan Gao is the Chief Technology Officer of Mavrx.
In this interview, Mavrx’s Yuan Gao explains how instead of investing in their own drone fleets, most businesses will look to a new type of service provider.
Executives are beginning to think about how they can take advantage of new data captured by drones, such as detailed imagery of crops, infrastructure, or facilities. But instead of investing in their own drone fleets, most businesses will look to a new type of service provider that will perform both data collection and sophisticated analysis. To better understand how this new model works, I recently spoke with Chief Technology Officer Yuan Gao of Mavrx, a San-Francisco-based provider of agriculture services.
PwC: How are drones being used commercially today?
Yuan Gao: There are four major non-military uses of drones: sport, cinematography, surveillance and surveying. And then there are minor uses, such as delivery.
Sport is the use of drones for recreational purposes. You’ve seen the drone racing competitions popping up here and there, and drones that are specifically for racing.
Cinematography is anything to do with taking videos and photography. The drones at a minimum are able to carry cameras and fly in a stable manner.
Surveying is what we’re doing, which is using drones to comprehensively cover large areas of land to map out certain things that you want to look at. In agriculture, that’s mapping out help using special cameras. But there are also things like mining, for example. In certain kinds of mining, you want to know how your mine has changed. You can do a 3-D mapping using a drone. Or in forestry, there are all sorts of use of surveying.
Surveillance is one of the more challenging uses. That involves being able to send a drone to a particular location to obtain persistent surveillance. And obviously that’s one of the more contentious issues these days, in terms of privacy. But there are other less objectionable to some people, uses such as security in a building site, for example.
PwC: What’s the main benefit of agricultural surveys of the kind you do?
Yuan Gao: The efficiency and environmental benefits are considerable. Let’s take the use of pesticides and other chemicals, for example. Most farms have a tendency to put down a lot more fertilizer and a lot more pesticides and insecticides than strictly needed. If you can determine exactly where in the field your fertilizers, insecticides, pesticides are needed, than that gives you an opportunity to target just those specific areas of the field. To improve your efficiency, you should be treating these fields as on a per foot or even per inch basis.
Being able to take a sensor and image the entire field and then work out exactly where to put down those chemicals will save you money. It’s best for the environment, keeps the chemicals on, and the farm should continue to increase in efficiency. When you can use them, drones give you better insights into field conditions than satellite imagery can.
Drones have revolutionized “scalping,” the process where a farmer or an agronomist would go into the field to inspect crops for problems and to determine how far along they are in their growth cycle. If there are eight farmers in a group, then all eight would fan into the field at the same time. The drone imagery saves them exponential time and greatly enhances accuracy. They can pinpoint where to go. We may not be able to say exactly what the problem is, but we can definitely alert farmers to deviations that need examining. So that saves them time and catches everything.
So that’s the benefit to the farmers directly.
PwC: How does that compare to what farmers rely on today—imagery from satellites? Is there a delay with satellite imagery?
Yuan Gao: The delay varies from 24 hours to several weeks, depending on which satellite you use and who’s operating it. For farming, you want to have the data within a few days, and you don’t want to have it two weeks down the road, things can change by then. But I think satellite companies and satellite services are currently operating within that time range.
The problem is if there’s a cloud, then you’re not getting that data. So if your satellite only comes overhead once every week, and on that particular day at that particular time there’s a cloud overhead, then you’re not getting that image. And farming tends to happen in places where there is a good amount of rain. And places with a good amount of rain tend to have a good amount of clouds. That’s one of the problems.
As for the resolution, it’s somewhere around 30 centimeters. The US government has a limit on what can be sold commercially. So even if satellite companies do collect higher than that, they’re not allowed to sell it. And that limit was reduced from ½ a meter to around 30 centimeters sometime last year.
PwC: Are you only using drones for your surveys?
Yuan Gao: Drones are not necessary for everything. They are a piece of the puzzle. Drone systems and automation only make sense if the human that we are replacing is a significant piece of the cost. For example, you wouldn’t necessarily want to automate an oceangoing cargo ship because the cost of the labor is insignificant compared to the value of the cargo it’s carrying. So eliminating headcount is only going to save you a small amount of money. Whereas with surveillance, you are actually able to reduce your cost by quite a large margin by increasing the amount of area that can be covered with security.
So drones don’t always make sense. In fact, that’s something we found first hand, and that’s exactly why we’re using manned aircraft today–because they cost less.
PwC: Why should farmers hire a service instead of doing the surveying themselves?
Yuan Gao: Data collection and analysis isn’t something that’s very easy for farmers to do. If every farmer needed to go work out how to hire a plane and fly over that field, that’s going to take them a lot of time and effort. So in most cases, farmers don’t like doing that.
What we’ve seen before is that there are small companies dotted across the U.S. who offer survey services to farmers. And so they take on quite a large amount of area in terms of acreage. And therefore they can benefit from the economy of scale. Instead driving out to one place and deploying a drone and then coming back, they can send a manned aircraft and cover quite a lot of area. Much more area than a small drone could cover.
So if drones are to replace this kind of activity, they need to be either so cheap and autonomous, so prevalent that every farmer can have one and be able to use one with minimal training. Or they need to be big, long-range drones that can cover a large amount of acreage.
Ultimately, a drone will be better than a human pilot can, because the flying the plane cover an area of land and just going back and forth, back and forth, covering it in a tight pattern. Covering everything over an acre of land. And that takes a long time; it takes a lot of concentration. And if you’ve ever been in one of these flights it’s very nauseating. It’s not a fun thing. And so in that respect, the drone would be better. But they would need to fulfill this scale requirement first.
What type of drone-collected data could your business benefit from? Let us know in the Comments.