Slashdot Mirror
Air Force Wants Technology That Will Let Drones Sense and Avoid Other Aircraft
coondoggie writes "With an eye toward letting drones share the nation's common airspace, the Air Force has set out to find the technology that will let unmanned aircraft sense and avoid other airplanes in flight. The ability to sense and avoid — common on all manned aircraft that fly the national airspace — is one of the trickier issues for drones which do not support such technology. It will be a major hurdle to jump as drone vendors and others press for common drone access to national airspace."
It is that easy? and a good script to read the output
Sometimes it's better not having signature
Why don't they invent technology that lets plane sense and avoid drones?
....it would probably be cleaver enough to also hit another aircraft. All they would need to do is change an equals to a not equals.
like a sheep dog!
They want to use unmanned drones to spy on everyone but right now that would be too dangerous because of possible collisions with commercial aircrafts.
Some kind of thing or person that controls air traffic
and also they could have some kind of thing or person that navigates the sky
IF ONLY THERE WERE SOLUTIONS TO THESE PROBLEMS
You made a Warner Bros cartoon featuring Wile E. Coyote come to mind, with the copter doing a sudden jump up and all the drones crashing nose-first midair.
Laughter is the Spackle of the Soul.
Defense contractors: But boss, I don't understand. We've been working 30 years and we've spent 100 billion dollars to get this thing to home in and hit any target you designate with a high probability, and now you want us to design one that does exactly the opposite? I suppose I could point you to one of our earlier test models...
Seven puppies were harmed during the making of this post.
ward, you were awfully hard on the beaver.
My units cost about $2 million each with about $100K maintenance annually, however maintenance facilities can range from 1-3K sq ft each. Throw in a about $500K for an ejection system to salvage my units in case of an emergency, you're looking at $2.5M up front per drone.
I call my prototype a PIE LOOT.
I swear to God...I swear to God! That is NOT how you treat your human!
They'll avoid other aircraft if they're not in the air
Asimov's robotic laws would be a great baseline for governing the behavior of these robots... err... UAV's.
ward, you were awfully hard on the beaver.
That's what she said,
blissfully.
I'm a guy who's been in the electronics and software industry for over 40 years and have a very strong background in RF, digital and analog systems.
I recently started putting that expertise to work in an attempt to come up with an effective and affordable "Sense And Avoid" (SAA) system -- at least in part because I fly FPV RC model aircraft and for these to be flown safely "beyond visual line of sight", some form of SAA is required.
After spending a considerable amount of time investigating previous strategies and considering the strengths and weaknesses of the available technologies, I have designed and prototyped a system that delivers a 1-mile "sphere of awareness" around any craft on which it is installed.
It does not rely on transponders (thus will "sense" *any* potential threat within that 1-mile sphere) and is small/light (250g) enough to be fitted to all but the smallest unmanned craft. The price (in volume production) would also be very reasonable -- about US$250.
Initial (ground-based, static) testing has shown that the prototype system conforms very closely to the design goals and expectations -- the next step is to strap a second prototype to a small foam RC model plane and start collecting dynamic data which will be used to test and refine the firmware.
Unfortunately -- this is where everything turns to custard.
The national airspace administrator here in New Zealand is CAA (our equivalent of the USA's FAA). They, in their infinite wisdom, have decided that since what I'm working on has significant commercial potential, I can not continue my development work (ie: strap this thing to a small foam RC model and fly it over a grassy paddock in the countryside) without first gaining a "commercial operating authority".
Now I've been flying RC models for almost 50 years and have a very high level of skill. Hell, I have two very popular YouTube channels with a total of over 45 million views and 100K subscribers in which I entertain and inform folks on the subject of RC models. However, all this counts for nothing and, according to CAA, if I want to continue my development of this technology by strapping it (as a passive payload) to the type of small foam model that thousands of folk fly here every weekend, I must jump a raft of ridiculous hurdles.
Firstly, the "minimum requirement" is a full-sized pilot's license -- which costs about $18K to obtain in this country.
Secondly, I have to file all sorts of safety plans, obtain a radio qualification and engage in a huge amount of bureaucratic crap -- simply to do what I've done as a hobbyist for decades -- fly a tiny (900g) foam RC plane over a grassy field in the countryside.
Now I don't have $18K to spend getting a pilot's license, besides which, this is silly bureaucratic nonsense!
As a result, the technology which I've developed and which stands to be a real "game changer" with massive export/earnings potential for this tiny nation that keeps crowing about its "innovative tech sector" is becalmed because some idiot desk-jockies seem to think that somehow, simply because what I'm doing has commercial potential, any RC flying I do will result in widespread death and destruction -- unless I spend months filling in forms, learning to fly a full-sized plane and licking boots.
This, my friends, is why New Zealand barely qualifies as a first-world country and will *never* play any significant role in the tech world.
Meanwhile, the same country spends $1m of taxpayers' money on something as lame and dangerous as the Martin Jetpack.
Go figure!
Those who ask "why not just find a quiet spot and test it anyway without telling anyone?"... well CAA have advised me that if I dare to do this without the required "authority", they will take "enforcement action" against me. So, if I turn around and say "I've tested it and it works" then it's "do not collect $200, do not pass go, go directly to jail".
And for those who ask "if this technology works as well as you say, why not get
Keep drones out of U.S. airspace. If this is done, there can be no crashes of drones into manned aircraft. As a no-cost benefit, the government won't be able to use drones to spy on citizens.
Circle the wagons and fire inward. Entropy increases without bounds.
So, basically TCAS? Add the required responder to the drone and the TCAS implementation and it'll know where all other nearby planes/drones are as well as when one is on a collision course.
See: http://en.wikipedia.org/wiki/Traffic_collision_avoidance_system
Site & blog: http://www.mayaposch.com
The technology has been in use since the 50's. How hard can it be to replace the robotic voice with a signal to the drone's navigation system?
It's just not mature yet for a broad solution. Some have achieved for a single platform, in a specific scenario, but it's really the first iteration before a solution can be applied in an environment like the national airspace. This is being worked on in both strategic level systems like the GlobalHawk and tactical levels like the Shadow and FireScout, and solutions range from ADS-B transponders, to active LIDAR.
FTA:
Given the results of Research and Development efforts over the past few years, it is now possible to equip [a drone] with technology that will address some of the major requirements currently driving the FAA authorization process.
The transponder route seems to be the lowest-hanging fruit at the time.The FAA will have to make ADS-B mandatory across the board so nobody in a Cessna ends up with one of those annoying quad copters through their windshields. At that point one of these might be a good starting point for every Joe Amateur Drone Dude wanting to fly with the big dogs:
http://www.sagetechcorp.com/unmanned-solutions/unmanned-solutions.cfm
http://www.unmannedsystemstechnology.com/2013/01/uav-sense-and-avoid-capabilities-demonstrated-by-sagetech-and-ing-robotic-aviation/
Just like everything these days; plenty of possibilities and approaches, not enough research dollars to get them to market. Plus all the design by committee dynamics that makes any solution take 10 times longer than anyone imagines.
"Now, I doubt any of you would prefer a rolled up newspaper as a weapon against a dictator or a criminal intruder."
You mean we currently have drones flying around, many of them outside of active warzones and over US cities, that *don't* currently have the availability to detect and avoid other aircraft??
If the masses can keep you down, you're not the Ubermensch.
like a sheep dog!
If safeguards are not put in place to prevent this, yes, especially if operator not-in-the-loop collision avoidance algorithms are used.
Speaking as a somewhat lapsed pilot, sense and avoid on manned aircraft is actually pretty poor. Sure, you know where to look if ATC tells you where you are in the circuit or to look for traffic, but there are a few little known facts (to the public anyway) that all converging traffic appears stationary to both pilots, a/c on the same altitude appear on the horizon to each other, and the horizon is where the greatest contrast change is in daylight (ground=dark, sky=light). In short, its pretty damn hard to spot the thing that isn't moving w.r.t. you and is on a cluttered poorly lit horizon. If it was easy, TCAS wouldn't be necessary for large commercial a/c.
Sense and avoid requirements have been difficult to establish for UAVs since Human sense and avoid capabilities really aren't well understood to begin with. If the only requirement were to be 'as good as' humans, we'd have the job pretty much done...'cause humans are so bad at it. Its only the traffic rules and ATC that keeps things ticking along relatively incident free. The danger from UAVs in NAS is really about reliability and flight over populated areas, but with this improving all the time, the only real barrier is the Sense and Avoid boondoggle. Its worth solving as a last minute safety, sure, but it provides a lot less security than the good old aviation regulations and good airmanship.
In August 2011 there was a midair between those aircraft. Thankfully no one died. Anyone know how that happened? Hopefully whatever solution is being proposed would have prevented that.
https://en.wikipedia.org/wiki/Automatic_dependent_surveillance-broadcast
There is also other existing tech, such as TCAS and TCAD but ADS-B will be a considerable
improvement over what exists now.
The ability to sense and avoid — common on all manned aircraft that fly the national airspace — is one of the trickier issues for drones which do not support such technology.
It's common all manned aircraft? But you can't stick the same thing in a drone? Why not, exactly?
this signature has been removed due to a DMCA takedown notice
I'm putting an aircraft beacon in my yard. YOU KIDS STOP FLYING OVER MY LAWN!
Radar is commonly use for this function ;]
Jack of all trades,master of none
See and avoid also means flocks of birds. I have hit birds in flight. One or two is bad enough. I have successfully avoided flocks of birds on takeoff that would have been very difficult to avoid with a drone.
This isn't trivial. Airliners have gone down because of it.
UAV, "Marco!" Other aircraft, "Polo!"
How many years have drones been flying for DoD? And they are just now getting around to thinking about how to avoid other aircraft? Is that "air superiority" or "air-rogance" or just stupidity?
Can you test on a kite?
Seriously,
You're too damn good to follow the rules. That's your problem. You think that the rules shouldn't apply to you because you're smarter than them. Really? You're smarter than 50 years of experience well studied and codified? You're so damn good that you can go fly your airplane without the same training that everybody else who shares the airspace has? You're embody the real problem with drones; people who don't want to pay for safety. It's $18K for a pilots license. Learn how to fly safely. Or, take your prototype and put it on a real, safely designed airplane, and go fly it with a real pilot.
Yes, I am a drone pilot. Yes, I've used the roughly $100K of training that's been put into my skill set to keep people alive. No, I don't want people like you every flying over my children. If you want to play with your toys, play by the toy rules. If you want to fly commercially, learn to fly commercialy, but get the fuck out of the airspace until you do.
Oh by the way, 1 mile range, beak to beak, gives less than 15 seconds against high speed intruders *flying legally*. Your approach is cute, has been considered by the adults, and discarded because it's irrelevant.
So one set of the software goes on the airplane, and the other goes on the missiles.