Unless I'm mistaken I feel that diydrones would rather all flights use a Tx, but are we not very close to a point where we can to put an end of the Tx for drone multirotors as the primary flying device. After all this is 'diydrones' not 'diyrc' and a Tx is just not very drone like.
I know there are some bad a$$ flyers out there, but for me I want more of a drone and less of an RC.
Push a button, take off, do a mission and return back home and autoland, use virtual controls if needed, more drone like.
With all the new functionality from that Arthur Benemann has packed into the latest droid planner 2, including 1 button take offs, follow me, dronie, guided and large screen phones and tablets it looks like we may be close. All the strides that DJI and Parrot has done with virtual joysticks that are extremely accurate cant we get away from requiring the Tx.
We have geofence RTL, battery low RTL and tons of failsafes, we have continue mission if lost signal for those long missions.
The Tx is a great backup to the phone, but I have no backup for my Tx anyhow so why would I need a backup for my phone/tablet. If battery on phone dies it can return launch.
Somehow the new Ghost is not even requiring a Tx at all.
Just my thoughts as id like to say bye forever of the requirement of a Tx :-)
Replies
I tried that about a year ago, strapped a beam on a medium-size Hexacopter and mounted a prop/motor on it, as a pusher, to see what would happen. It didn't work well, I think because when pushing it tried to tilt the aircraft, and the FC then essentially pushed back to try to keep it level so instead of zooming around and remaining flat it moved only very slowly.Maybe what it needed was to have the push happening at the center of gravity, so it wouldn't induce tilt. Fun to play with, but I didn't take it any further.
Hi Chim, thanks for your answer, very interesting.
Was the pusher motor being driven by the autopilot or was it governed externally?
Sorry for the slow reply. Pusher motor was controlled directly by the pilot via a dedicated channel and knob on the Tx. So it was totally independent of the APM.
Hi Randy,
Thanks for the reply, thats great news. We are in the days of phones and tablets that are stable, fast cpu's, loads of memory, long battery life and highly accurate retina displays. With the amazing work you guys have done with the latest firmware and Mr. Benemann making the Droid Planner 2 I would think adding accurate virtual sticks to Benemanns app should be feasible.
The virtual joysticks could be a slide up from the bottom that takes a large portion of the screen for more accurate flying not just micro flying, even take full screen on newer versions of iOS and Android.
Another nice thing about a tablet that is using a ground module over your traditional Tx is that my Tx has a finite amount of functions I can set with switches, tablets can send a virtually unlimited amount of commands from a ground module.
Gary, yeah prob not the APM then :-)
Ed,
We've had this discussion about virtual joysticks before (about 2 years ago iirc) and they simply don't give the haptic feedback required for accurate flight control. Sure, you can develop quite good sensitivity with using them, but you lack the signal your brain needs (while it's looking at the aircraft) to know the position of the 'sticks' relative to neutral and thus the control problem is much, much harder.
Some have suggested that you use the internal accelerometers on the phone and attitude/rate estimation as an input signal for control... but unfortunately the accelerometers in the phone tend to be far lower quality than those in the IMU and prone to a lot of noise, making accurate state estimation difficult. Sure, there's more processing power available on the phone to run better estimators, but that also introduces more delay into the control loop, again, making the control problem harder.
The easiest solution is dedicated hardware for fine grain generation of the pilots input signal, that provides suitable haptic feedback and a range of functions to enable it to control a large range of craft... and that is currently (and has been for a long time) the R/C transmitter. If you can create something that has all of the functionality of this device, you've still only recreated something that already exists. I cannot see a phone or tablet doing this, or incorporating additional hardware to support this, simply because the market isn't there.
When we get to accurate voice control or gesture-based input, then maybe it'll be time to ditch the Tx... but we're a LONG way from that (in spite of the few lab demonstrations trying to demonstate progress in that direction).
Cheers,
Tim
@Tim: Excellent explanation of the fatal lack of feedback from "virtual" as opposed to physical joysticks. I've flown both, the latter extensively, and the former quite a bit, and can verify the effect.
@ Edgar: IMHO it is irresponsible and dangerous to launch any aircraft of any appreciable size (say over about a kilo) without being able to competently and comfortably take off, fly tail-in and nose-in, hover in place decently in various conditions, fly circuits in both directions, maintain altitude, re- establish lost orientation, and land within a couple of feet of a given spot without drama, all manually.
People who cannot do this but fly anyway place themselves and everyone else at risk and are a general threat to the entire drone community. Autopilot systems as they stand today are wonderful but they, and/or the things they depend upon (GPS, for example), fail much too often to be trusted completely, as just about anyone with a couple of year's experience knows. Most people who don't fly manually have no idea just how quickly a serious quad that gets a bad command can literally be in someone's face.
Vendors/makers would hate this, but maybe what's needed is firmware that requires a demonstration of manual flight before it allows autonomous functions to be enabled. People who refuse to learn to fly should confine themselves to ground-based robotics, or at most to very small aircraft, period. That may change, but we are not nearly as close as a lot of people hereabouts like to think.
As for getting rid of trad Tx's, aside from the haptic feedback matter there's an issue of reliability, notwithstanding Mr. Scott's assertions that tablets/phones etc. can be depended upon. Tablets and phones, (not to mention laptops, etc.) in fact crash all too often, at least that's been my experience which goes back to PDA's and such. They are inherently subject to internal and external interference, all sorts of bugs, they're underpowered, easy to drop, and on and on.
Traditional R/C transmitters on the other hand, these days even more inexpensive ones, are extremely reliable. There's nothing there but a very basic radio transmitter, some simple switches, and two joysticks - all things that are completely mature in terms of technology and production, zero mysteries with the bugs chased out long, long ago.
So, the idea of abandoning the trad Tx and thereby avoiding having to learn to fly is a fail-fail. It fails because learning to fly remains, or should remain, for good reason, mandatory. And it fails because it suggests replacing stone-reliable powerful hardware with this week's overpriced stylish twinkly little offerings from Best Buy or a company named after a fruit. Phones and fondle-slabs will replace real Tx's about the same time they replace real cameras.
@Chim: Well said
Much better than my earlier knee-jerk reaction.
I am by trade, an IT Developer of said laptops, tablets and smartphones, so the issues with bugs, crashes and hacking are things I am all too well aware of and, in fact, I do relate to the frustrations using a Tx.
However, a typical airline's 747 has even more sophisticated autopilot and a glass cockpit among it's other systems, but they are careful and thoughtfully segregated from all the other systems on board.
Smartphones and tablets have every function using one interface, one CPU and one antennae. The result is important and unimportant tasks have to -share- and in some cases (yes) compete for the same resource.
So if you get a phone call while your flying, the device automatically takes away CPU and resources from your flight, just to decide whether or not to deal with it. There are also concerns about the potential for viruses jumping on board causing more problems.
A separate Tx has none of those concerns and so will remain so.
Yes you can fly a drone from a tablet and for toys they work well to introduce someone to the skill needed for flying.
But on that 747, the stick remains and the pilots are guided to continue their self-discipline and self-reliance in piloting aircraft. One hopes that if those electronics go offline, the stick and the pilot's skill can still put it down safely.
A contest between two philosophies, easy of use vs security.
You -can- make it really easy to use, but if your tech fails, just how secure/safe is your flying?
Are you going to be responsible for the operation of your Apple smartphone flying your drone, when that company won't even let you pop the back-cover off to change the battery?
It may be moot in the end.
The FAA is currently recommending pilot's licenses for commercial drone piloting. Unless you want to spend 10x more on the pilot's license than what it costs for the drone itself (now, a Cessna actually costs about as much as a low-end ($13K) car to build, its the -other- costs that make it add up to more than $100K), you may end up hiring out for a drone pilot afterall.
Thanks Tim, love your feedback on haptic of a physical stick. And your right, virtual is not as accurate or give the feedback. But allowing flight without requiring one to bind to a Tx should on the radar as I feel we have came a long way in the past 2 years and don't feel that we are that long of a ways away. Of course we need to have enough manual control for a safe landing or fly it back is needed is all would be required, not to have granular control as say someone say sport flying.
There are a few solutions I have seen
DJI solution, basic virtual joysticks
Ghost that "seems" from what I read a way to "guide" the aircraft quad during land and flight.
And lastly the Parrot solution one that would cover haptic feedback but would require hardware
Although they use a proprietary device I would think a solution could be made that used a phone/tablet using the micro usb and an app like DroidPlanner. Parrots also commonly use sonar to help landing.
The day will come where we we won't require a transmitter but before then we need to increase the reliability so that it is close to a human pilot's. Each release gets safer but we still have a few situations where the autopilot can't recover on it's own, the worst being the rapid climb caused by very high vibrations on the copter. AC3.3 will have the EKF turned on by default and optical flow should also work very well so that will be a big step forward in reliability.
AC 3.3 is for Pixhawk up only is it not Randy? I don't think it will be coming to 2.6
-
2
-
3
-
4
-
5
-
6
of 6 Next