Shenyang/Hongdu Li Jian ("Sharp Sword") UCAV prototype

[Skipjack]

Well right now they are too big for small fighter jets. The future may be different.

The future is nearly now. For example, F-35B was designed so that the shaft-driven lift fan could be replaced with a generator that powers directed-energy weapons.

But then, some drones right now are not that much smaller than manned fighters.

Because they need to carry those big satellite dishes and complex, unjammable, encrypted, real-time data links to do anything useful.

They save money however, because of the things needed for supporting and protecting a human life are missing.

You forgot to add the cost of onboard, real-time data link equipment, expensive comm sat channels, ground-based control and uplink/downlink equipment in Nevada, and replacement costs due to their high crash and capture rates.

Drones Most Accident-Prone U.S. Air Force Craft

... Northrop’s Global Hawk and General Atomics’s Predator and Reaper unmanned aerial vehicles have had a combined 9.31 accidents for every 100,000 hours of flying. That’s the highest rate of any category of aircraft and more than triple the fleet-wide average of 3.03, according to military data compiled by Bloomberg.

USAF Splashes One Reaper

Operators lost control over the unmanned aircraft during its operation. With the UAV headed in a direction where it was about to depart Afghanistan's air space, a U.S. Air Force aircraft brought down the Reaper in what the Air Force says was a remote part of Afghanistan.

Iran–U.S. RQ-170 incident

On 4 December 2011, an American Lockheed Martin RQ-170 Sentinel unmanned aerial vehicle (UAV) was captured by Iranian forces near the city of Kashmar in northeastern Iran. The Iranian government announced that the UAV was brought down by its cyberwarfare unit which commandeered the aircraft and safely landed it, after initial reports from Western news sources inaccurately claimed that it had been "shot down". The United States government claims that the UAV malfunctioned and crashed.

Air Force Strategic Choices and Budget Priorities Brief at the Pentagon, January 27, 2012

GEN. SCHWARTZ:
The bottom line on your multiple questions -- let me start first with the rationale. It was our expectation, our -- certainly our hope, that the advantages that a Global Hawk-like platform provides would -- which we anticipated both would be cost of operation, on the one hand, and clearly persistence on the other -- would play out in practice.

The reality is that the Global Hawk system has proven not to be less expensive to operate than the U-2. And in many respects, the Global Hawk Block 30 system is not as capable from a sensor point of view, as is the U-2. And so we have made the choice, as the deputy secretary mentioned yesterday -- cancel the Block 30 program.

Finally, having drones that are as capable (or even almost as capable) as manned fighter jets, means that you can employ completely new tactics.

I have never argued that drones and manned fighters would not be used in combination. I have argued against your assertion that drones can replace manned fighters within a few decades.

Electronics will get cheaper and so will the electronics used by the UAVs. They will also get smaller. Dont forget that we have only just begun this path. In a few decades things will look completely different.
Of course drones crash more often. They are still suffering from problems compared to manned fighters. But when they crash, no pilots get killed and that is what counts.
My argument was that the development of new manend fighters will be less important as drones take over some missions and reduce the risks for the manned fighters.
It makes more sense to improve the weapons, detection and system capabilities of the manned fighters to make them more efficient and focus the development on new drones.

[djolds1]

Do some googling on combat flight dynamics, nonlinear systems theory and how far machines have yet to go to match the broadband processing power, sensor fusion, inventiveness, intuition and experience of one well-trained human brain.

It will be decades, at best, before a UCAV can defeat a top-of-the-line fighter flown by a capable pilot. Pilots are currently trained to evade "AI" SAMs/AAMs, which can pull many more gees than a larger future UCAV.

Note that I am not saying that UCAVs can never reach that level of dogfighting.

"AI" as programmed into today's munitions is nothing near to what the Watson engine has already demonstrated. It may well take decades to develop and deploy such drones, but that will be an issue of dysfunctional budgeting and acquisitions programs. All necessary software and hardware already exists in the lab. Slow integration is a matter of politics, not capability.

I don't think you really appreciate how Watson has changed the game wrt "AI." Do some reading.

By the time that capability is attained by UCAVs, however, compact, speed-of-light, directed-energy weapons and distributed-aperature sensors will nullify their high-gee advantage, and their small size will become a disadvantage.

That is a separate issue. Solid state lasers are progressing nicely, but remain massive and so limited to Surface Combatants or fixed installations. OTOH, you are correct that eventually air dominance will be reduced to Hammerverse norms. But absent fundamental breakthroughs, shrinking those solid state lasers will not be a rapid matter.

For a plausible fictional description of tactical laser installations, John Lumpkin's "The Human Reach" series is first rate. Read both books in nearly one sitting.

Wrong. And getting wronger every day. Not even counting game-changing advances in HTSCs, LENRs, metamaterial optics, etc.

Ironic. You're so near-term techno-optimist on some subjects, yet so techno-pessimist on others.

[ladajo]

We are not far off at all on widespread deployment of lasers for air and limited surface defense. Watch what happens in the next year or two.

The ship board UAV Swatter is already out on maiden T&E live deployment. I posit they will propagate across units in a similar model to how Remote-Op Sabilized guns and Flir balls were. Shared kits, and then eventually permanent installs as funding and yard availabilty became resourced.

There is also plenty of hardware work being done for airframe deployable Swatters and mess with other flying things units. Google around.

Before you know it, there will be a number of driving, floating and flying units with laser shooters on them. Althouhg it looks like floating will take the lead again.

[djolds1]

We are not far off at all on widespread deployment of lasers for air and limited surface defense. Watch what happens in the next year or two.

The ship board UAV Swatter is already out on maiden T&E live deployment. I posit they will propagate across units in a similar model to how Remote-Op Sabilized guns and Flir balls were. Shared kits, and then eventually permanent installs as funding and yard availabilty became resourced.

Something the size of a 5" Naval turret may fit into a 747 like the ABL of a decade past, but it is not suitable to slot into fighter aircraft - which is precisely what DeltaV was waxing poetic on upthread.

[DeltaV]

"AI" as programmed into today's munitions is nothing near to what the Watson engine has already demonstrated.

Good. AAMs/SAMs are supposed to destroy enemy aircraft, not win at Jeopardy. Tell the engineers at, say, Raytheon that their AAMs have no "smarts" regarding their mission. Do you really think such algorithms would be in the public domain? And yet, even the best AAMs/SAMs can be outsmarted by a properly-trained human.

All necessary software and hardware already exists in the lab.

Citation needed.

I don't think you really appreciate how Watson has changed the game wrt "AI." Do some reading.

What is it with your Watson fixation? Do some reading.
http://en.wikipedia.org/wiki/Watson_(computer)

Watson is a Question answering (QA) computing system built by IBM. IBM describes it as "an application of advanced Natural Language Processing, Information Retrieval, Knowledge Representation and Reasoning, and Machine Learning technologies to the field of open domain question answering" which is "built on IBM's DeepQA technology for hypothesis generation, massive evidence gathering, analysis, and scoring."

http://en.wikipedia.org/wiki/Question_answering

Question Answering (QA) is a computer science discipline within the fields of information retrieval and natural language processing (NLP), which is concerned with building systems that automatically answer questions posed by humans in a natural language.

Well, fine, if a Watson-based UCAV wants to hold a conversation. Maybe it will frustrate the opposing fighter pilot to death by emulating phone tech support from India.
No, when UCAVs eventually match humans in a dogfight it will likely be some combination of neural/fuzzy algorithms with background genetic-algorithm optimization, possibly mixed with digital/analog optical/SAW processing and, maybe, quantum computing.

Solid state lasers are progressing nicely, but remain massive and so limited to Surface Combatants or fixed installations... But absent fundamental breakthroughs, shrinking those solid state lasers will not be a rapid matter.

Wrong.

As of 11 years ago:
Lasers Being Developed For F-35 and AC-130
Lasers, HPM Weapons Near Operational Status

More recently:
High Energy Liquid Laser Area Defense System (HELLADS) Programme

HELLADS is a diode-pumped, liquid-cooled solid-state laser weaponry system. The system will be ten times lighter and compact in comparison to other similar high-power laser systems.

The order of magnitude will facilitate integration of the system with aircraft, ground vehicles and UAVs. It is expected to have a maximum weight of 750kg and volume of two to three cubic metres.

The system is being developed with a power to weight ratio of about 5kg/kW.

[hanelyp]

One thing drones can't do yet is identify valid targets by themselves. That will require a Watson level or better intelligence.

One thing drones should be very good at is reacting Very Fast to changing sensor data, potentially running through the OODA (Observe, Orient, Decide, Act) loop several times for each time a human pilot does it once. As I hear it, a faster OODA loop can win more dogfights than making the right decision more often.

Anyone else recall a game called CRobots? It was a kind of video game you played by writing a control program for a simplistic on screen drone. The best programs I wrote were horribly sloppy in their aim, but reacted very fast.

[ladajo]

We are not far off at all on widespread deployment of lasers for air and limited surface defense. Watch what happens in the next year or two.

The ship board UAV Swatter is already out on maiden T&E live deployment. I posit they will propagate across units in a similar model to how Remote-Op Sabilized guns and Flir balls were. Shared kits, and then eventually permanent installs as funding and yard availabilty became resourced.

Something the size of a 5" Naval turret may fit into a 747 like the ABL of a decade past, but it is not suitable to slot into fighter aircraft - which is precisely what DeltaV was waxing poetic on upthread.

The actaul current systems are smaller than 5 inch turret size. You are looking at environmental packaging. Also recall that we are discussing hard kill systems. If you want to talk about soft kill systems, they are even much smaller, and flight worthy in tactical aircraft.

As I hear it, a faster OODA loop can win more dogfights than making the right decision more often.

There is more too it than that. Boyd's OODA intent was more than just thinking faster thatn the other guy. It was also about keeping him off balance, tkaing and keeping initiative, amplifying his confusion. To this end, one of the things Boyd advocated was radical random action in the mix as well.

[ladajo]

Give this a read...

https://opencrs.com/document/R41526/2012-10-19/

[Betruger]

Boyd is really fun to read about...

[djolds1]

I don't think you really appreciate how Watson has changed the game wrt "AI." Do some reading.

What is it with your Watson fixation? Do some reading.
http://en.wikipedia.org/wiki/Watson_(computer)

Watson provides a synthetic option to fields typically credited as requiring human intuition - like say piloting combat aircraft. This is being used in medical diagnosis, not simply conversation on game shows - natural language is how the Watson engine absorbs info, but does not limit how it can in turn apply that info.

The order of magnitude will facilitate integration of the system with aircraft, ground vehicles and UAVs. It is expected to have a maximum weight of 750kg and volume of two to three cubic metres.

The system is being developed with a power to weight ratio of about 5kg/kW.

After review it appears you're correct. I concede the point.

The actual current systems are smaller than 5 inch turret size. You are looking at environmental packaging. Also recall that we are discussing hard kill systems. If you want to talk about soft kill systems, they are even much smaller, and flight worthy in tactical aircraft.

Good point.

[Skipjack]

That does not change the fact that the same lasers that defeat UAVs would also defeat manned fighters. Combat UAVs are not THAT small. Also, I presume that the lasers that fit onto a fighther jet can not fire unlimited amounts of times without service or a period of recharge.
Also remember that pilots are more valuable (and cost more ) than the planes.

[ladajo]

Also remember that pilots are more valuable (and cost more ) than the planes.

That is a point of discussion in itself. But if it is going to happen, someone needs to agree on the foundation blocks of "value" first.
In any event, I think this entire discussion is missing the point of how air combat is done, and folks are mixing apples, oranges and eggplants.

Let's try and sort that first.

Air Combat is three major regimes and several sub-regimes.
Air to Air - This includes two sub arenas of Counter Protection and Counter Strike
Air to Ground - This includes many sub regimes, but can be boiled into essentially Over Land and Over Water. An example of the sub-regimes is target sets and Fires concepts, as well as soft kill, hard kill, direct strike, stand-off, etc.
Air Situational Awareness - This is about knowing what is up and where, as well as purposes. It is an entire animal to itself.

Unmanned, just like manned vehicles can find effective roles in any of these regimes. But, yes there are certain specific things they are not yet ready for prime time. The oft argued Dogfight is one of them. That said, they are still useful for Air v. Air, if you understand how the fights are prosecuted, if you do not, you will want to argue the "whites of their eyes", which is not in general terms correct.

Another way to to look at the Air Combat problem is by "Operational Function" areas. The standard bins are Maneuver, Fires, Sustainment, Command and Control, Intelligence and Force Protection. If you have not idea what I am talking about then try reading this (And I picked the Army version on purpose, if anyone wishes I can post the Joint version or other service, as they all read the same - which is my point here):

usarmy.vo.llnwd.net/e2/rv5.../ADP_3-0_ULO_Oct_2011_APD.pdf‎

Every thing that flys fills a role in at least one functional area, and possibly more. The point being, what flys has a purpose and nitch. There is no such thing as a true multi-purpose aircraft and for that matter, ship, soldier, etc. I think in this discussion to date (and many dates prior) has generally gotten lost in that false idea. Drones, just like manned aircraft will and do fill required nitches. As technology advances, more nitches can be filled, just like with manned aircraft. One of the funny points I find here is the idea that air to air combat is done primarily by dogfighting. Not so much boys and girls. It is all about the tried and true american way of war principle, "shoot them with a big bullet when they as far away as possible, before they can shoot at you." As romantic as "dogfighting" appears and is painted by Hollywood, it is just not the way. Dogfighting will get you dead. As noted before, neither planes nor people are cheap. That is why we have weapons, you win a war by exhausting the other guy's resources and will. If you are tossing out weapons you can not afford, that cost magnitudes more than the system they defeat, then you are generally in the wrong over the long fight.

Now we can get on to the real argument.

The main argument I am making here is that war costs National Blood, Treasure and Will. He who wants to win must have an objective that is commensurate with his ability to chase it down within the means he has, against an enemy that can not exhaust him in either will or treasure. You can make your opportunity, but there is always risk. This is where the warfare context term Assymetric came in to being. The idea was always there, just not so often in such extreme terms. Your own Risk is is increased due to loss of High Value Blood, Treasure and Will, by an extremely Low Risk Action with an extremely Low Value Blood, Treasure and Will expenditure. An exagerated imbalance. In modern warfare, Drones and Unmmaned vehicles have introduced a huge opportunity to pursue and effect this concept. And at a minimum, significantly reduce the imbalance in Blood, Treasure and Will. Drones and unmanned vehicles, properly managed in implementation have huge potential in return on investment. That is the point. You do not need to pitch a drone into a dogfight, it costs too much, and for now, may not win. It has perfectly good 'mini-drones' it can toss out (missiles) that can make the dogfight for it, just like manned aircraft have and use. Evading missles is a distractor. You may be evading the first wave, but when the second wave (augmented by a number of unmanned vehicles) shows up while you are in the middle of that...and it was in relative terms 'cheap' to get them into the fight compared with flying an equivalent number of manned airframes...blah blah blah...

[DeltaV]

One of the funny points I find here is the idea that air to air combat is done primarily by dogfighting. Not so much boys and girls. It is all about the tried and true american way of war principle, "shoot them with a big bullet when they as far away as possible, before they can shoot at you." As romantic as "dogfighting" appears and is painted by Hollywood, it is just not the way.

You are parroting the same old song that led to surprise in Vietnam:

U.S. went into Vietnam relying on AIM-7 as main air-to-air missile
–
Pre-war estimated Pk: 0.70
–
Demonstrated Pk: 0.08
–
MiG 100 times likelier to make it to close fight than expected

BVR has improved somewhat in the AMRAAM era. Since 1991, 20 of 61 kills have been BVR. One third.

Over-reliance on BVR is stupid:
Promise and Reality: Beyond Visual Range (BVR) Air-To-Air Combat
by
Lt Col Patrick Higby, USAF

This paper has shown that the pursuit of costly BVR capabilities during the Cold War was not justified by actual BVR performance. Air-to-air combat has not transformed into a long-range slugfest of technology wherein radar-guided missiles score near-guaranteed kills. Human factors, such as pilot skill—or the opponent’s ineptness—still trump technology. Furthermore, BVR appears to work best in situations it is needed least.

[Skipjack]

Also remember that pilots are more valuable (and cost more ) than the planes.

The main argument I am making here is that war costs National Blood, Treasure and Will. He who wants to win must have an objective that is commensurate with his ability to chase it down within the means he has, against an enemy that can not exhaust him in either will or treasure. You can make your opportunity, but there is always risk. This is where the warfare context term Assymetric came in to being. The idea was always there, just not so often in such extreme terms. Your own Risk is is increased due to loss of High Value Blood, Treasure and Will, by an extremely Low Risk Action with an extremely Low Value Blood, Treasure and Will expenditure. An exagerated imbalance. In modern warfare, Drones and Unmmaned vehicles have introduced a huge opportunity to pursue and effect this concept. And at a minimum, significantly reduce the imbalance in Blood, Treasure and Will. Drones and unmanned vehicles, properly managed in implementation have huge potential in return on investment. That is the point. You do not need to pitch a drone into a dogfight, it costs too much, and for now, may not win. It has perfectly good 'mini-drones' it can toss out (missiles) that can make the dogfight for it, just like manned aircraft have and use. Evading missles is a distractor. You may be evading the first wave, but when the second wave (augmented by a number of unmanned vehicles) shows up while you are in the middle of that...and it was in relative terms 'cheap' to get them into the fight compared with flying an equivalent number of manned airframes...blah blah blah...

That is exactly what I am saying as well.

[ladajo]

One of the funny points I find here is the idea that air to air combat is done primarily by dogfighting. Not so much boys and girls. It is all about the tried and true american way of war principle, "shoot them with a big bullet when they as far away as possible, before they can shoot at you." As romantic as "dogfighting" appears and is painted by Hollywood, it is just not the way.

You are parroting the same old song that led to surprise in Vietnam:

U.S. went into Vietnam relying on AIM-7 as main air-to-air missile
–
Pre-war estimated Pk: 0.70
–
Demonstrated Pk: 0.08
–
MiG 100 times likelier to make it to close fight than expected

BVR has improved somewhat in the AMRAAM era. Since 1991, 20 of 61 kills have been BVR. One third.

Over-reliance on BVR is stupid:
Promise and Reality: Beyond Visual Range (BVR) Air-To-Air Combat
by
Lt Col Patrick Higby, USAF

This paper has shown that the pursuit of costly BVR capabilities during the Cold War was not justified by actual BVR performance. Air-to-air combat has not transformed into a long-range slugfest of technology wherein radar-guided missiles score near-guaranteed kills. Human factors, such as pilot skill—or the opponent’s ineptness—still trump technology. Furthermore, BVR appears to work best in situations it is needed least.

BVR, VR and Gun Range are all very different animals.

BVR = Typically Radar and other tech to gain terminal ranges.
VR = Radar, Visual, and other tech to gain terminal ranges
Gun = Radar, Visual, and other tech that must BE AT TERMINAL ranges.

Nobody disagrees that shooting a BVR at a target that can see it coming is a low .pk event. However that said, there are some pretty smart weapons today in the BVR arena, vice the Dumb ones of days gone by. They are called fire and forget for a reason. Sparrow, Phoenix were never fire and forget nor meant to be.
The entire Radar Missile BVR don't work argument is based on fire and continue support to terminal weapons. AMRAAM and follow-on are much better for a number of reasons. Unfortunately, Post Vietnam US systems do not have a lot of real combat data to access effectiveness. But the data there is indicates magnitudes improvement in missile systems. Since Vietnam, our opponents have choosen for the most part not to fly against us. Those few that did, did not normally get home. How many of those were gun kills? In the Gulf, it was ZERO for 33 kills. I think the last time anyone was killed with guns was in the Ethiopia/Eritrea dust up.

Even the new variants of Sidewinder are magnitudes better than those from The Days of Yor. Your argument is dated and irrelevant when discussing modern systems. Yes, if you wish to discuss Cold War aquisitions and theory, unless some one does not know what they are talking about, they will fully agree that there was a pre-mature emphasis given to Air to Air Missile engagements. Nowadays, the missile systems are ENTIRELY different animals. You can not mix the old with the new in this discussion.

Stuart Nichols wrote a decent paper back in the late 90's discussing this very argument. Please note that a decade and a half has elapsed in technology since he did this work. I recommend you read it.

www.dtic.mil/cgi-bin/GetTRDoc?AD=ADA398674