Unmanned Cargo Aircraft??
#91
Banned
Joined APC: Feb 2017
Posts: 2,275
Well, lets see... Fewer pilots needed, erases any leverage we have, reduced pay, furloughs, basically the destruction of our livlihoods, and ability to provide for our families. Every CEO, in every industry is trying their best to reduce the need for enployees. The killing off of the middle class.
#92
No bucks, no Buck Robot
Well, lets see... Fewer pilots needed, erases any leverage we have, reduced pay, furloughs, basically the destruction of our livlihoods, and ability to provide for our families. Every CEO, in every industry is trying their best to reduce the need for enployees. The killing off of the middle class.
#93
On Reserve
Joined APC: Mar 2017
Posts: 18
Well, lets see... Fewer pilots needed, erases any leverage we have, reduced pay, furloughs, basically the destruction of our livlihoods, and ability to provide for our families. Every CEO, in every industry is trying their best to reduce the need for enployees. The killing off of the middle class.
#94
Banned
Joined APC: Feb 2017
Posts: 2,275
#95
On Reserve
Joined APC: Mar 2017
Posts: 18
Holes can be cut on both sides of the argument. Anyway all the best, hope your predictions on the matter don't come to affect you and others
#97
1. See and Avoid. The sensors aren't there. On my current airplane we have the best weather radar you can buy for an airliner and still it can be fooled or attenuated. Whereas, the CA and I visually see some clouds and determine "no it's not safe to fly thru there."
Do we have systems that can anticipate problems? For example, I'm landing on a runway, an airplane is approaching a hold short rather fast, and I see that the airplane isn't stopping and I initiate a go-around. If it is possible, why don't we have it today to prevent runway incursions?
2. Compound Emergencies. Qantas Flight 32 suffered an engine failure which generated over 100 ECAM messages! A logic ladder system? Really? There's no way a programmer is going to be able to think of every possible scenario and all of its variables to be able to prioritize checklists and decision making. The Qantas crew had decades of experience and they used it to determine what was the correct next step or decision to be made. There's no way a programmer can possibly program something this complex and all of the different scenarios, variables, or outcomes.
Another example, in a recent LOE simulator event, we were presented with a seemingly simple problem. We had two EICAS checklists to run. The problem, both checklists conflicted with one another. We had to use our best judgement as to which checklist we were going to override and which one we would execute.
3. Computer/Signal Hacking. The USAF has lost HUNDREDS of drones. Their reliability is shocking. No way could you accept that for commercial aviation, manned or unmanned. But that's ok because we don't have humans on them and we're flying them mostly in very sterile airspace over war zones.
However, the hacking, jamming, and spoofing of navigation systems or datalinks can be done with cheap commercially available products. This is a huge vulnerability and it has not been solved. Anything is hackable. The question is are we going to allow potentially hundreds of these UAVs get airborne with the possibility of being jammed (or whatever) and they fallback on whatever the programmer programmed in such an event? Could we plan for something like that?
4. Risk versus Reward. Right now it is simply too expensive with a lot of risk to do pilotless airplanes. Until some of these major issues are solved and preventable, it just can't happen anytime soon.
The only way it will happen sooner is if some billionaire or major corporation is able to buy off Congress and bypass safety system redundancies and regulations written in decades of blood shed by aviators who died on the job.
#98
Don't you think that creating a pilotless surface vessel would be easier? Why is no one looking into that?
I have extensive experience with unmanned aircraft (policy) in the Department of Defense, and liaison with the FAA, and I can assure you, this is not even close to happening.
Ask yourself why we don't have single engine freighters.
This is a fantasy of cargo operators who don't view pilots as human beings, and begrudge them every nickel they pay them.
I have extensive experience with unmanned aircraft (policy) in the Department of Defense, and liaison with the FAA, and I can assure you, this is not even close to happening.
Ask yourself why we don't have single engine freighters.
This is a fantasy of cargo operators who don't view pilots as human beings, and begrudge them every nickel they pay them.
#99
Banned
Joined APC: Feb 2007
Posts: 461
Don't you think that creating a pilotless surface vessel would be easier? Why is no one looking into that?
I have extensive experience with unmanned aircraft (policy) in the Department of Defense, and liaison with the FAA, and I can assure you, this is not even close to happening.
Ask yourself why we don't have single engine freighters.
This is a fantasy of cargo operators who don't view pilots as human beings, and begrudge them every nickel they pay them.
I have extensive experience with unmanned aircraft (policy) in the Department of Defense, and liaison with the FAA, and I can assure you, this is not even close to happening.
Ask yourself why we don't have single engine freighters.
This is a fantasy of cargo operators who don't view pilots as human beings, and begrudge them every nickel they pay them.
And there are single engine freighters. (ship and airplane)
I can't tell if you're being sarcastic or not.
#100
Line Holder
Joined APC: Feb 2015
Position: FO
Posts: 36
If you've ever witnessed USAF unmanned ops, you'll recognize the argument behind his first statement ("several technical challenges that must be solved before airlines or cargo freighters fly in the same airspace with the rest of us"). The USAF is as-good-or-better at remotely piloted ops than anyone else out there. That said, when they launch or recover a SINGLE RPA it closes down the associated runway for 15-20 mins. Imagine that happening at an LAX or JFK...it would be a disaster. So the first challenge to solve is that one...namely we either:
A) Build a totally separate infrastructure of airports exclusively for RPA's to operate in, or
B) Somehow get much, much better at the launch and recovery than the AF has been able to do in 15 years.
Assuming we can solve that problem, PTB's next few points deal with the complexity of non-normals, or EP's...whatever you want to call them. Many flying training organizations teach a set of heuristics (rules of thumb) to deal with EP's. The one I learned was:
1) Maintain Aircraft Control
2) Analyze the Situation
3) Take the Appropriate Action
4) Land as soon as Conditions Permit
another common one is:
Aviate, Navigate, Communicate
Sets of rules like this work great for a human mind, because it has the adaptability to take a simple rule (like maintain aircraft control) and, with some training, quickly reach a solution that generally yields a safe outcome. But lets stop to think about how you would program something as simple as "Aviate" in a procedural programming language.
Function AVIATE
- Inputs: airspeed, altitude (MSL and AGL), attitude, heading, AOA, Thrust setting, G load, flight control configuration
- Output: Primary and Secondary flight control inputs, Thrust commands
Big picture...what procedural commands do you program? Lets try..."Maintain course, altitude, and speed".
What if the current course and speed is 2000 fpm down at 500' AGL? What if the current course and speed is 20 deg NH at FL 350 with airspeed rapidly decreasing? So maybe we program something like "Don't hit the ground, don't stall, otherwise maintain course and speed".
What if we're flying in the vicinity of tallish mountains, and we don't have adequate terrain clearance 20 miles ahead on our current course and speed? Should be covered by our previous command: "Don't hit the ground". But how? Climb to get over it? What if we've lost an engine? "Fly the single engine departure procedure". What if there isn't one at this airport for some reason. Should we turn left or right?
But even assuming we can program "Don't hit the ground, don't stall, don't fly into a mountain, otherwise maintain course and speed"...this isn't always the right answer. What if a non-normal happens on final approach? Sometimes the right answer is to Go-Around...sometimes the right answer is to land. If I was teaching a human, I'd tell him "If on final with an EP, and able to make a safe landing, then continue, land, and figure it out on the ground." But what if (as PTB was implying) the non-normal is the proverbial bus full of nun's pulling onto the runway. If the runway is long enough, you can probably make a safe landing beyond the bus, but is this what we want to program? Maybe so, if we also have a second non-normal...like Cargo Fire. The point is the procedural code for this would be outrageous...
if Bus Full of Nuns = TRUE & Cargo Fire = FALSE
-- Thrust = GA, Pitch = 15 NH
else if Bus Full of Nuns = TRUE & Cargo Fire = TRUE
-- Land the airplane beyond the nuns
else Bus Full of Nuns = FALSE
-- Land the airplane normally
Maybe I have an overly simplistic view of the limits of programming at present, if so I apologize. But I still think his point is valid: Trying to build and test systems that can operate in the 121 environment at the level of safety and reliability required, without human oversight is (right now) not possible.
Thread
Thread Starter
Forum
Replies
Last Post