Slashdot Mirror
FAA Adds a Study On Adding Drones To Commercial Aviation
coondoggie writes "Facing a number of technical challenges, the Federal Aviation Administration said today it added another research project designed to better understand how unmanned aircraft can be brought safely into the national airspace. The FAA set a two-year research and development agreement with Insitu (an independent subsidiary of Boeing) and the New Jersey Air National Guard that will help FAA scientists to study and better understand unmanned aircraft design, construction, and features. Researchers will also look at the differences in how an air traffic controller would manage an unmanned aircraft vs. a manned aircraft."
Only now they're not quite so goddamn funny.
Dog is my co-pilot.
But..but...why would our government want to spy on its own citizens???
I swear to God...I swear to God! That is NOT how you treat your human!
"Why are you so against this military hardware used against our enemies? It's not like the government will be flying these things over its own citizens."
Fast forward a few years....
A few important points about this:
1. They are not talking about autonomous UAVs. These UAVs are essentially remote-controlled aircraft piloted by real pilots. I think some people assume these things think for themselves but that's not the case. Now that doesn't automatically discount concerns of safety, but "skynet" is not the case here.
2. This is not specifically for military only. Many uses for UAVs exist outside of military applications such as basic transport. Of course they'll use them for surveillance, but they already do that with aircraft. UAVs can simply linger longer because one pilot can take over during flight. Similar to how large aircraft do it now with redundant crew members.
UPS and FedEx and other air cargo type things I could see as a huge advantage.
Eventually refining the confidence and quality of the AI to the point where it could haul actual passengers. I'd bet that they mean time between failures of machines could out pace that of human error fairly quickly so it'd actually be safer.
Remember the Elevator had the same type of history. There was a time when an attendant was there to push the button for you as a way to reassure everyone that it was safe. Eventually people learned they could push the button on their own.
Paying taxes to buy civilization is like paying a hooker to buy love.
Don't call my ass "nothing", you insensitive clod!
This isn't the FAA building and deploying UAV's on any kind of scale. This is the FAA trying to figure out how to safely integrate UAV's into the national aerospace system (NAS). Personally, as a pilot, while I distrust the FAA to some extent, as the agency charged with ensuring safety of all operators in the NAS, they are the right agency to be performing this study.
When some other agency says they're going to start launching UAV's in the NAS, the FAA needs to have ammunition to enforce safety measures to ensure that the UAV's not pose an undue hazard to other aircraft and that the UAV operators respond accordingly to instructions from air traffic control.
Merde, il pleut encore!
Already happens: http://www.drugwarrant.com/articles/drug-war-victim/
Comment removed based on user account deletion
Skynet jokes aside, drones are both useful and inevitable. And not only the winged ones. Look for a possible resurgence of blimps and airships in widespread use. Hang a radar on a blimp, park it at high altitude, and you have an instant radar system upgrade for air traffic control. Or for border patrol. Or for search and rescue. Etc etc etc. The uses for UAV's in the civilian sector are endless.
Life is hard, and the world is cruel
Thing is, military drones have no people on board. Passenger jets would have people on board.
Why would they do it? it's all about saving money, it's not in the interests of passengers.
So, do you think Sully would have pulled off a perfect water landing if he had been miles away from the cockpit? If the pilot's life isn't at risk, I just don't think he's going to have the same drive to handle an emergency. He's not going to have all the visual, auditory, and tactile, information a human in the pilot's seat is going to have either. Sometimes you need the reflexes of a well trained human being whose life is on the line.
Give me Classic Slashdot or give me death!
And we can't be certain if knowing that you will surely survive any catastrophe is not actually at least as strong deterrent & motivation - after all, you know you will face the consequences if that was your fault.
And what about the possibility of the remote pilot ditching and making a run for it? At least with the local pilot they'd need to parachute out of the plane. In this case, Jim takes a coffee break and never comes back.
So, do you think Sully would have pulled off a perfect water landing if he had been miles away from the cockpit?
I don't see why not. I don't know if I'd be under any more or less stress to handle an emergency if my life is on the line or if its hundreds of passengers.
And essentially, some of the mistakes that might cause emergencies will be reduced by a drone that doesn't forget things. And for the record, no human has ever had faster REFLEXES than a modern computer, we've just had the gift of INSIGHT. The only reason we can outsmart a computer in various fields is being able to do what it does not expect. Usually a pilot in the sky is best for situations where your target is also human so you need to be able to have someone who can continue another person's train of thought based on one action (Like fighter pilots watching how enemy pilots fly, and Rescue teams saving victims). It's the unexpected elements of humans that require another human to assist.
In the case of a commercial airliner needing to make an emergency landing, I don't see why a computer couldn't handle it, should proper emergency protocols be handled. The only problem I see arising is instrumentation error, which a few cameras in the cockpit could fix.
"The third and by far most complex step (ERAM Release 1) is the replacement of the Host Computer System with new software and hardware ... national deployment begins in FY 2009 and concludes in FY 2011"
That's all well and good except for the part about it not lasting more than 6 days when they tried to use it in production. They may be *trying* to replace the old system. Whether they're succeeding at replacing it is a whole other question.
This could be much safer, actually. Granted, something like choosing the best place for a crash landing may not be the kind of thing you could so easily program, since it would require information such as where there were likely to be people, what the ground seems to be made of and such. You can't very easily pre-program this since you don't know where a plane is going to be needing to make an emergency landing. On the other hand, you can have this sort of thing raise a flag in a remote location. There you can have the best pilot in the world on stand-by to pilot such crash landings remotely, and he can be someone specially trained to do nothing but crash landings, and since he isn't on the plane, you won't lose him in those situations that can't be saved. Today you are stuck with whatever quality pilot happens to be in the cockpit, and having experience with crashes is likely to mean that the pilot is dead and his experience wasted.
So if the remote control technology can be made secure enough from tampering and reliable enough, then this sort of thing could even make crashes less deadly. Imagine having a Sully in every plane that crashes.
Let me preface my comments by saying I'm actually on one of the teams working on this problem at the FAA Tech Center, and that I was actually there during the signing ceremony yesterday (though I was mostly just annoyed by the loud music coming from the lobby interrupting my work and turned down the opportunity to take a bus ride to go see the ScanEagle fly). While we're excited they're loaning us two ScanEagles, we're already pretty deep into studying this problem. My group is on our third (or is it fourth?) simulation study right now, and we're ramping up for a gigantic one that will study a mix of GA, Commercial and about four to six UAS systems in a mixed-use airspace around the January timeframe. So now that we've gotten that out of the way, I'd like to address some of your comments. :)
Actually there are significant differences between controlling a UAS and a normal aircraft for an ATC. First and foremost is that it's a pilot's responsibility to see after the safety of his or her plane at all times, even if that means disobeying a directive from ATC. UAS pilots just aren't capable of this. They literally can't look out the window and see if they're going to run into someone or something. They also may be controlling more than one system at a time, which is something you never have to worry about in a "normal" ATC scenario. On top of this it seems that UAS operators may or may not be instrumented rated, which means they're not always trained for flying in controlled civilian airspace.
Oh and then there's the fact there are still many areas in US Airspace that have no radar coverage whatsoever. Yes, I'm serious. ATC depends on pilot reports in those areas. In the future we'll be practically eliminating radar for en-route ATC environments in favor of a satellite based solution like ADS-B.
And the final difference which the article actually touches on is that what we're simulating is a future airspace: one where we've moved on to trajectory-based operations. Currently an aircraft in controlled airspace moves along airways (or jetways), which are like one-lane highways that go from one point to another. This is why some people always see a lot of flights over their house, there's one or more airways crossing over it. Air traffic control's main responsibility is to make sure no aircraft comes within a certain distance of any other aircraft along these airways (usually five miles in an en-route environment, though that can vary based on conditions and aircraft types). At the traffic levels we're expecting in 10-25 years, this system breaks down.
In the future, aircraft will fly more direct routes to their destination (instead of navigating a graph of nodes via one-way connections), and ATC will be modeling their trajectories in four dimensions to make sure those safety bubbles are magically maintained. We haven't even finished figuring out how this will work for manned aircraft, much less when you add UAS to the mix (though a UAS will probably keep to their plotted flight plan better than a manned aircraft, but I digress). That many UASes and manned aircraft sharing airspace and traveling in all different directions is a scary thought right now, but that's why we're being so serious about simulating these things and developing very strict procedures for how it will be run.
In most cases, though, with a modern fly-by-wire system the pilot has significantly better situational awareness, response time, and contro
Maxim: People cannot follow directions.
Increases in truth directly with the length of time spent explaining them
Would a drone using forward looking infrared even have hit the geese in the first place though?