Yamaha Demos Agricultural Drone, But Humans Can’t Unleash It Yet

from IEEE Spectrum

Agriculture is (arguably) where drones are going to be the most useful in the near future. Drones have already proven their usefulness to hobbyists and the military, and agriculture is the next huge commercial frontier. And although commercial drone operation is still waiting for official approval, at least in the United States, developers of unmanned aerial vehicles are eager to show that the technology is ready.

Yesterday, Yamaha and UC Davis put on a demonstration of the RMAX unmanned helicopter in Napa Valley, in the heart of California’s wine country. Yamaha’s RMAX is basically a scaled-up version of a nitro (or electric) hobby sport helicopter, or a scaled-down version of a real helicopter.

It’s beastly enough to carry a 28-kilogram payload, and Yamaha has it set up to spray vineyards. It’s complex and expensive, but the company is making a case that it’s the most efficient and cost effective way to treat high value crops, especially with an eye towards autonomy.

When we say that the RMAX can’t be “unleashed” yet, we’re talking about autonomy, or the lack thereof. Currently, the helicopter communicates via a radio link to a controller operated by a human, and it is unable to fly autonomously. There’s an autonomous component to the flight control software, in that the helicopter can hold its altitude, speed, and heading, or hover in place without a human having to manage it. But that’s it. The rest of the time (all of the time, essentially), a human is working a controller, flying the helicopter back and forth over the vineyard rows entirely by hand.

This is not ideal for a whole bunch of reasons. First, it’s not something that humans are particularly efficient at. We can do it, but since the pilot is flying the drone while standing in one spot, they’ve got just one viewpoint on what’s going on, which makes it difficult to fly a spraying pattern without building in a lot of needless overlap, wasting both time and money and adding more chemicals that necessary.

Second, these drones aren’t messing around: they’re very large, very scary vehicles, and are difficult to control (even with practice) due to both the inherent instability of all helicopters and the fact that while the controls of a helicopter operate from its perspective, a remote pilot has to mentally compensate for any angles of yaw whenever the helicopter isn’t facing directly away from them. So, if something goes wrong (mechanical failure, gust of wind, bird strike), an autonomous system will likely be better suited to getting the drone back down to the ground safely.

And third, lack of autonomy has a significant impact on the cost effectiveness of the system. To fly it, Yamaha sends out a team of three people. There’s a pilot, a safety observer to help the pilot fly in the right places and not crash into anything, and an assistant who deals with refilling tanks and moving stuff around. With an autonomous system, on the other hand, you could likely reduce that to one person: a safety pilot who could manage the hardware aspects, and then stand by to take over manual control if it ever became necessary.

Yamaha, though, is sticking with the humans in the loop system for the foreseeable future. They have enough trouble getting clearance from the U.S. Federal Aviation Administration for this totally human-controlled version of the drone. The FAA has very good reasons for being as cautious as it is, and we don’t want to suggest that the agency should suddenly lift all bans on drones. Having said that, government regulation is the biggest obstacle to deploying the RMAX in the United States, autonomous or not. It’s been operating successfully in Australia for years, and in Japan since 1991, seeding and spraying rice crops.

Talking to viticulturalists from UC Davis, on the other hand, made things seem a little more optimistic for the near future of autonomous agriculture. UC Davis is partnering with Yamaha to evaluate the RMAX on a research vineyard, so they’ve had plenty of experience (in additional effort and frustration) jumping through all the necessary FAA hoops.

UC Davis is actively working towards autonomy for things like spot spraying, where the drone sprays only in the very specific areas of a vineyard that need it, identified from (say) multispectral aerial maps. Again, this would save time and money while minimizing chemical use. Another cool thing that you could do with an autonomous helicopter is to identify areas in a vineyard susceptible to freezing at night, and then send in the helicopter to use its rotor wash to keep the plants moving and thawed out.

Putting aside the issue of autonomy (for now), do very expensive (low six figures) drones really make sense in an agricultural context? Napa Valley may be a special case because aerial spraying using traditional cropduster aircraft is prohibited, which means that spraying is usually done by tractors driving along each row. The RMAX can cover twelve acres in the time that it takes for a tractor to cover one, and that’s on flat ground. Where Yamaha foresees the real benefits are on hillside applications, where it’s dangerous (and sometimes impossible) for ground vehicles to operate effectively. The RMAX doesn’t care in the least whether it’s hovering over flat ground or a steep slope, and spends the same amount of time covering hillside crops.

Continue Reading at Spectrum.IEEE.org…