Why not unmanned fighters? -duck-

Hi All,

Please understand that I'm not looking to start a flame war. Just an honest question from an interested party. Here goes...


After the F-22 some were predicting that it may be the last manned "super fighter" ever. Why? Because...

• the planes of today/tommorrow are capable of far more than our frail human bodies are.
• there are design effeciencies that come from eliminating space, displays, controls, etc for the pilot
• today's best airframes are already fly-by-wire so why not fly them remotely

So... with that premise in mind, there must be some reasons for NOT replacing certain roles with remotely piloted vehicles. What are they?


(Again, please... no flames. Trying to learn here)

One though I have always had regarding unmanned flight is the possibility of the systems being hacked while in flight. There have already been claims that drones have been captured and system hacking has been suspected.

One thing about having a pilot in the cockpit is that you can't hack a pilot.

Well, Bullet is the expert on this topic around here, but I'll provide a couple nuggets for thought.

RPAs require an encrypted up-link to their user. Drone sends encrypted update to user via satellite, user decodes and enters new command, command is encrypted, sent via satellite and decoded by the drone. This all takes times - a few seconds. In a dogfight, milliseconds matter and you can't break the speed of light! If your answer is "well, automate them," then you enter a whole new ballgame. Who is accountable and responsible when they attack (enemy or otherwise)? How automated can they be? How much code and development (a LOT) would be able to allow them autonomy for fighting?

There is much to be done before we can realistically do this.

Thanks for the thoughts so far, folks.

As a follow up on the reaction-time thought... I suppose my ignorance is showing through here. And I know the "death of dogfighting" has been predicted many times before. But with today's first-see, first-shoot capabilities and stealth technologies... don't those and the ability to physically perform manuevers that are simply impossible for a human body to withstand offset some/all of the minor lag?

MedB said:

Thanks for the thoughts so far, folks.

As a follow up on the reaction-time thought... I suppose my ignorance is showing through here. And I know the "death of dogfighting" has been predicted many times before. But with today's first-see, first-shoot capabilities and stealth technologies... don't those and the ability to physically perform manuevers that are simply impossible for a human body to withstand offset some/all of the minor lag?

Click to expand...

The same thing was said about the F-4 in Vietnam until ROE prevented many of the longer-ranged kills.

Actually, there is an excellent book out "Wired for War" (and I cannot remember the name of the author) that looks at the whole remotely-piloted-robotic concept from mini-tanks to aircraft that is extremely interesting. It addresses the very concept of when you break the human link and when you should not.

Other than that, almost all air warfare requires a positive identification of your target and that usually boils down to an eyeball connected to a human brain. The German and British nightfighters both had to have a human ID before they shot during WWII, the airwar over North Vietnam required it as noted above, and the Gulf wars required it as well. In fact, it seems like the longer the range you can acquire a target, the more you have to have a positive ID. Get a crowded airspace (admittedly with good mix of RPAs) and the problem gets really big fast. Americans don't like fratricide. Another soft spot is the actual data link itself. As sophisticated as it may be in carrying data, it is still a jamable radio signal and subject to the laws of physics.

Simple answer that others have a piece of...

If you are going to pilot it remotely, even in a simulator with perfect super HD 3-D vision in surround, we don't have the bandwidth to present that much data without the delay would make his/her reaction too late.

If we put the decision compute and execute (i.e. pilot decision and action) in a computer on-board, the flying data center would not fit in a fighter. The pattern matching, ability to change engagement rules dynamically, learning capability found in the live, trained human brain takes racks of servers and storage.

Engaging a host of identified and potential enemy combatants while trying to appropriately engage and destroy only the planned and opportunistic targets without taking out your own, flying a highly-engineered machine pushing its engineering limits in physical environment with unpredictable atmospheric challenges is a supercomputing exercise if I have ever heard of one. I have worked in a facility with a supercomputer. They are quite big.

MedB said:

. . . So... with that premise in mind, there must be some reasons for NOT replacing certain roles with remotely piloted vehicles. What are they?


(Again, please... no flames. Trying to learn here)

Click to expand...

The idea of replacing certain roles with remotely piloted vehicles will always end up "flaming" discussions. Rather, we need to consider changing how we operate, not "replace."

A big consideration is preventing fratricides. Can we conduct an air operation where only RPVs will attack the enemy?

If one F22 cost $50 million and a RPV cost $5 million, can we change how we conduct certain air operations since we can send 10 RPVs instead of only one airplane with human pilot.

Can we do different mission planning as we won't send a human pilot on a suicide mission but we can send RPVs on suicide missions?

MemberLG said:

The idea of replacing certain roles with remotely piloted vehicles will always end up "flaming" discussions. Rather, we need to consider changing how we operate, not "replace."

A big consideration is preventing fratricides. Can we conduct an air operation where only RPVs will attack the enemy?

If one F22 cost $50 million and a RPV cost $5 million, can we change how we conduct certain air operations since we can send 10 RPVs instead of only one airplane with human pilot.

Can we do different mission planning as we won't send a human pilot on a suicide mission but we can send RPVs on suicide missions?

Click to expand...

USAF says F22=$140+million. GAO says more.

I'm no expert, but I think that RPVs just don't give you the same situational awareness, and perhaps the larger issue is the ethical questions that arise once you start having machines killing people instead of people killing people. Just some food for thought.

How is drone-to-air any different than drone-to-ground?

Also, do pilots today actually "see" the enemy before hitting the fire button?

because ground targets arent going at 800 plus miles per hour, you just don't have the situation awareness.

Data-link issues (lots of complex technology and stuff...)
Situational Awareness issues (A lot of RPA operators said it was "like looking through a straw," which would hurt in a dogfight!)
R&D/production costs (they would be BIG)

Thanks for the education!

Getting the sense that the tech is not quite there yet to support it (cost, bandwidth, security). Makes sense.

My son's goal has been to become a PJ (CRO). It's one of those roles that you hope for many reasons he never gets to put into practice! Maybe by that time there will be at least a few less folks in harms way due to remote-piloted craft.

Thanks again,

MedB

Another perspective on command time. Comm satellites are generally in geosynchronous (more likely geostationary) orbit at ~35,000 km altitude. The time it takes for a signal to reach it and back (speed of light) is 0.25 seconds. Without autonomy, you're talking about a 0.5 second lag minimum in strict signal travel time for the drone to speak to the operator and the operator to send a command back. If you shorten the distance with a large satellite network at a lower orbit, you introduce relay time (which actually would add more time if it has to use more than one satellite). A solution is to have the operator much closer to the drone location to reduce signal lag time. You still have encryption/decryption time problems on the signal. Frankly, the drone needs autonomy to be able to engage in air-to-air combat. And we're back to what the means in terms of development and legal.

Fun to talk about though and I second the endorsement of Wired for War.

Hi Hornet,

I guess that's where my neophyte perspective was headed too... why not co-locate the Remote Operators in an AWAC or some such right alongside the Battle Managers?

Or even... Let's put a couple of F-22s/other fighters out there along WITH the RP planes. The F-22 pilots control the RPAs directly whenever possible to stay at arm's length but still maintain awareness. But should it be needed, they fly in themselves. RPAs stay on station, F-22 pilots swap out as needed or some such.

Anyways... sure smarter people than me have thought of this.

MedB said:

Or even... Let's put a couple of F-22s/other fighters out there along WITH the RP planes. The F-22 pilots control the RPAs directly whenever possible to stay at arm's length but still maintain awareness. But should it be needed, they fly in themselves. RPAs stay on station, F-22 pilots swap out as needed or some such.

Anyways... sure smarter people than me have thought of this.

Click to expand...

And even smarter people will figure out how to get Congress to fund yet another advanced weapon system for a threat which may or may not exist 30 years in the future.

How's about we get the bugs out of the F-22, figure out how to get some help funding the F-35, and concentrate on fully exploiting the capabilities of the existing crop of drones and those in full development.

MedB said:

Thanks for the education!

Getting the sense that the tech is not quite there yet to support it (cost, bandwidth, security). Makes sense.

My son's goal has been to become a PJ (CRO). It's one of those roles that you hope for many reasons he never gets to put into practice! Maybe by that time there will be at least a few less folks in harms way due to remote-piloted craft.

Thanks again,

MedB

Click to expand...

Besides jets, there has been a lot of focus on Remote Piloted Helicopters. Manufacturers have made a goal to have working models by 2015.

To me , that's the scary part, the day they start trying to transport troops in aircraft that are flown remotely. As a PJ, I'm sure your son would feel a lot better knowing there was a human pilot in the cockpit of the plane or helicopter he just climbed into.

When it comes to technology, just because we can, doesn't always mean we should.

Well, a lot of the replies have already mentioned most of what I would have mentioned n this regard. Why not RPVs versus manned fighters? Let summarize:

- Technology simply isn't there yet. Bandwidth required, reaction time, network security, artifical intelligence required to see and react to the entire battlespace in a timely manner,... the list goes on. Getting better, but we are no way NEAR what Hollywood envisions what we would be capable of doing with RPVs.

- The air to air fight is the world's ultimate chess match. In 3-D. With closing velocities in excess of 1400 MPH. With missiles. And Jamming. A confusing, adrenaline pumping chaotic mess. And NO computer, on a network, with requirements for human apporval before hitting the pickle button, is able to react fast enough (yet). You can get around this somewhat by removing the human interface / approval. But that is how Skynet became aware, and we all know how that turned out.

- Are people looking into all the issues listed here? (who controls, RPVs being controlled int he air by F-22s / AWACs, networks, and all the rest). Simple answer: Yes. I will neither confirm nor deny that I am a part of that effort.

Hornet beat me to it and is quite accurate. There is NO LATENCY lower than the human brain. There can never be a time where a remote pilot, flying a drone, can have to reaction time as a live pilot in a plane. It's not possible based on the laws of math and physics. It's possible that technology can improve a remote pilot's ability to situational awareness to be that of a live pilot; but there will always be a delay between receiving information, processing that information, making a decision, sending that decision, and the drone executing that decision. That delay/latency can never be lowered to that of a live pilot.

However, it can be decided that such latency isn't important. I.e. In a bomber or recon, milliseconds are not as important. Some may even think that because the price of a drone is so much less than a real plane and pilot, that having a higher failure rate in a fighter scenario with drones is more acceptable.

My biggest problem with drones with weapons, especially fighter missions, is the desensitization that the Playstation pilots have. Even today, there are video games where you fly and kill using drones. These are games. The remote pilot, on the ground, 1500+ miles away, doesn't have nearly the concern that a live pilot would. You can NOT be shot down if you're a remote drone pilot. You're still going to see your spouse and kids come 5pm. If you miss your target and it hits something else because you were desensitized to the situation by being physically separated from it; you won't be as concerned. That's human nature. And the first time that an air to air missile is shot from a drone, and for some reason, the target turns out to not be what the Playstation pilot thought it was; shiite will hit the fan. Recon and ground to air, I can see. Not air to air.