|
| I'm experimenting with a few things:
Long-range autonomous missions.
For example, to fly to a "distant" pasture and record a video over the area. Since the drone is out of radio range, video is stored on-board and viewed after landing. Might be handy for doing a quick inventory of what's there, but the point of the experiment is to see how well the flight itself works. In my tests, it works well. The drone is flying a scripted mission, so it doesn't require pilot input in order to get along.
As for what it does in flight, that's open-ended. Missions can include a variety of commands (move a servo, take a picture, send a PWM value out an I/O port, read a sensor, etc), which means anything from dropping a payload to triggering a sensor is possible. It's a matter of building and attaching whatever "thing" you want to use to the drone, and connecting it (see below).
Limiting factors are lift capacity (function of battery power), flight time (battery power), and FAA rules requiring line-of-sight (LOS) flight. So when I say "long range", I'm talking about total mission length of maybe 4 miles. Batteries continue to improve, and flight time and lift capacity along with them. FAA rules will most likely adapt to allow long-range non-LOS flight, driven by the whole "drone delivery" hype in the news. (Probably not before sense-and-avoid tech becomes more widespread though.) For now, I'm out in the middle of big empty fields, testing the idea of letting a drone fly far away and essentially flying itself. It's worked well so far.
Unattended flight.
When you use an irrigator, you don't stand by the pivot and "drive" it, you tell it what you want and move on while the irrigator does what you told it to (ideally). If you could schedule or otherwise trigger drone flights, you'd be able to do something similar. I've built some rudimentary code to: 1) Generate a mission, 2) Send it to a powered-up drone, 3) Start the drone's motors ("arm" it), 4) Launch it on the mission, all without operator intervention. Those are the basics.
So, imagine you had some kind of GPS-savvy beacon on the end tower of an irrigator. The beacon would broadcast its location (say, across a mesh network) to a computer in your office. Any time the location had moved more than, say, 500 meters, the computer would generate a mission saying "fly to the pivot's lat/lng, start recording video, fly at 8 mph to the beacon's lat/lng at 60 ft, stop recording video at the end, fly home, land", send it to the drone, arm it, and send it on its way. Overall, the idea is that while you could still fly a drone for scouting, etc. you could also have drones working in the background on various jobs, just looking at their output when it's convenient. The flight controller I'm using even has code in it for controlling a sprayer from an autonomous mission. Obviously lift capacity/flight time is a limiting factor in getting a tank of pesticide off the ground and keeping it there, but maybe power systems will advance enough that automated spraying is feasible.
I haven't worked out how to park a drone somewhere and keep it powered up (so it can receive missions) while it's also being charged.
Attaching/connecting "things".
Tractors have 3-point hitches and hydraulic couplings, meaning there's a standard for connecting tractors to implements to get work done. No such standard exists for drones, although moves are being made. The 3DR Solo, for example, has an accessory bay on it that can be connected to the drone's internals and controlled either remotely (from an app running on a tablet attached to the controller), or as part of an autonomous mission. The spec is available for anyone to make use of, the idea being that people will make Solo accessories and market them. This is the "3-point-hitch/hydraulics" for a drone. It's still fairly new, so not a lot of accessories on the market yet. (So far, I know of someone developing a 100w LED for the Solo, turning it into a small remote-controlled sun. Which might be handy.)
Most drones contain just enough computing horsepower to fly, and very basic IO ports. So any accessories you attach have to be wired up to a specific IO port or something, and you're really limited in what you can do in terms of control. The Solo operates entirely on long-range wifi (range on mine is 1.5 miles), and it (and its controller) are both full-fledged computers. They're capable of talking to each other over a standard network, like normal computers do. Anything else on the Solo's wifi network can do the same. I'm working on setting up a BeagleBone (credit card-sized Linux computer with 96 IO ports) to attach sensors to, and run a web service on. I could attach that to the bottom of the Solo, send it a command to start gathering sensor data (whatever that might be) in flight, and have it all stored on an SD card when it returns.
Lots of possibilities, but it's still in the early stages.
| |
|
what can a drone do in regards to ag??
