Kaffis Mark V wrote:
Diamondeye wrote:
A tank or other major vehicle can't easily have the crew removed
Huh? Did I miss something? Does tank piloting put a greater demand and emphasis on low latency reaction times than flying? Google isn't just remote driving cars, they're AI-driving cars. NASA is remote/programmatically piloting drones on Mars. Obviously, the control inputs on a tank are different than a car, but the complexity and nature of the decision making isn't that different, nor the consequences of latency.
We could easily remote pilot tanks if we choose, I'm positive.
It's not as simple as all that.
First, the fact that Google is remote/AI driving cars doesn't mean that's going to turn out all that well if we tried to practice it with hundreds of thousands of cars on the road. It's one thing to make a few prototypes that you can put all your time and attention into making work, it's quite another when those prototypes have to work when produced at production-copy price. A number of weapon systems have been abandoned when this occurred; the prototype worked fine at $2,000,000 a copy, but when you had to start cutting the price down to $200,000 or $20,000 a copy.. things started not working so well. When I deployed, another officer in my unit was a civilian electrical engineer at one of the various Army arsenals.. masters, in EE, as a matter of fact, which was funny because outwardly he gave the impression of the biggest Alabama redneck you ever met. Anyhow, he had some stories about weapons systems that failed in spectacular and hilarious ways when tested with "production" specs. If you don't know how big a difference that is, take a look at pictures of the YF-22 compared the the production F-22. There's visible differences in the actual airframe, notably in the nose, wingtips, and horizontal stabilizers.
Now, moving on to your examples specifically:
First, you chose an excellent comparison, because a tank is probably the
best ground vehicle to think about automating. There's several reasons for this: They're already high on weight/power budget, they're a tracked vehicle, and their primary purpose is to directly identify and attack enemy targets with their main gun.
Simply making a tank able to drive remotely (lets leave AI aside for now) and identify and attack targets remotely could, no doubt, be done today. That would not be particularly difficult; remotely driveable vehicles are available as toys, and a full-size vehicle is essentially the same concept, just larger and much more complex. The tank already has a fully automated and stabilized fire-control system, so the basis for making the gunner operate it remotely is there. There's an internal crew compartment to put all the additional electronics in, so space and weight probably wouldn't be an issue, since we're taking out about 2,000 pounds of crewmembers, baggage, food, water, etc. In that sense, you're correct.
However, ground combat in any format (tank, artillery, infantry, engineer, etc.) is a lot more than just being able to drive around and shoot.
First, there's the situational awareness of the crew. An M1A2 has a crew of 4, every one of which has access to the outside world, and who will frequently operate "unbuttoned" with hatches open. This is for peripheral vision and general awareness.. but for the ground combatant, unlike the pilot, every sense except possibly taste has meaning and input. A pilot uses vision almost exclusively; he gets audio cues from the plane and talks on the radio, but he doesn't directly use hearing. He gets tactile feedback from the aircraft of how its flying, but that's it as far as touch, and he certainly doesn't smell much of anything. While short-range direct vision is important, the fighter pilot will rely mainly on his electronic sensors because the speeds and ranges he fights at mean that many targets are very hard to see or can't be seen at all. Tank crews use powerful sensors like thermal imagers as well, but they don't substitute for full, all-around awareness, and trying to electronically "pipe" this back to a bunch of tank operators is going to be very power and bandwidth-intensive to achieve similar quality to actually being there - and remember, you need to be able to do this for a lot more tanks than you would fighter aircraft.
Second, the crew itself. One crewmember can be eliminated right up front - the loader. This isn't that great a benefit though; other tanks have done exactly this, using a mechanical loader. The M1 varients have not, however, because mechanical loaders are not as reliable as 19-year-old privates. Remember that the loader isn't simply a mechanical device working in isolation; it's in an armored vehicle that bumps and jostles and jolts all over the place over open ground. This is hard on things. Simply removing the loader means there's more likelihood hood of ammo jams. That's ok; on a normal, 3- man crew tank someone can still presumably clear the jam. On our fully-remote tank, who does that? If this tank jams, it's now almost as useless as if it was destroyed.
This brings us to another problem - crew-level maintenance. Crew-conducted preventive maintenance and minor repairs are a major part of keeping equipment working on the battlefield. Combat vehicles are driven much harder, and under much rougher conditions than civilian vehicles typically are and require a great deal of maintenance. Maintenance management and discipline is a significant responsibility of mine as a commander, and relates to everything from monitoring fluid levels, to tightening bolts, to re-fitting thrown tracks on vehicles out in the field (yes, the crew typically does this themselves if the damage is not severe.) With no crew, how will this be done?
Now, back to the crew. We eliminated the loader. For the crew, we might be able to eliminate 2 more and have 1 guy operate it, but let's be conservative and say we need 2. Ok, we reduced crew needs by 50%. That's a good thing. We would do this by combining the commander and gunner into one position. Furthermore, we eliminate the need for command tanks at platoon, company, battalion, and brigade level since now the leader/commander can sit in his own automated control chair designed for command, and so can his platoon sergeant/XO/S3 (operations officer, at BN and higher level) without the need to fight their own tank. Those command tanks can then be consolidated into regular platoons.
Those are positives, but where is this control center? If it can be destroyed, or even electronically cut off from all its tanks.. you just knocked out an entire battalion in one swoop.
Now, let's move on to another subject - breaching. If you want to breach an obstacle, like a major minefield/tank ditch complex in front of a defensive position (breaching is a very complex and involved operation, and could easily support lengthy academic papers all on its own.. even individual aspects of it, such as fire support for a breach could) you need things like mine ploughs and rollers on your tanks. These are hard on the tank's frame, and slow it, so you also don't want them on any longer than necessary. How are they taken on and off if the tank crew is not there? Engineers and infantry? Ok, that could be done, but now you've increased their workload in an operation that is already going to be very physically intense.
Last, your examples - the Google car and Mars rover - have issues themselves. Mars rovers are like a once-every-5-to-10-years thing; Curiosity was launched in 2011 and the next mission is planned for 2020. Also, when it went into autonomous mode it was stationary.
CuriosityQuote:
In April and early May 2013, Curiosity went into an autonomous operation mode for approximately 25 days during Earth-Mars solar conjunction. During this time, the rover continued to monitor atmospheric and radiation data, but did not move on the Martian surface.[14][15]
Moreover, the Martian evironment, while tough, is nothing like combat. The only real threat is the terrain and maybe meteorites, and the latter probably can't really be dealt with anyhow. An entire team of scientists and engineers works on that
one vehicle, and everything on it could be custom-made for that mission since it was the only one of its kind. While ground combat vehicles clearly don't face the challenges of trying to get to Mars, they also need to be produced and operated in the hundreds or thousands, and thus the issues are totally different.
As for the Google car, the Google car is a little better (city streets are reasonably complex) but they don't have to deal with hostile action, and, as stated, are very much a prototype. The remote Google car is probably a much better candidate for comparison than a AI one, though.
Note also that aircraft fly for a few hours and then return to an airfield or aircraft carrier for full servicing, arming, and fueling; problems can't really be corrected in the air
at all. Ground forces need to be able to remain in action for weeks. Air and ground combat have fundamental differences so that what might be easy in one environment might be very hard in another. The same applies to the sea.
Not that making a remote tank is impossible, or won't ever be done. Maybe it will. It is, however, a long, long way off; to my knowledge it is not even seriously being looked at by anyone, for the reasons I cited. Any developer already knows they can make a vehicle drive and shoot by remote; it's all the other things that make up so much of combat and yet are not what people usually think of that make it impractical right now.
There's also the question of "what purpose does it serve?" Get fewer of our guys killed. Ok, that's legit, but our tankers are already very unlikely to take casualties compared to most other soldiers in contact with the enemy because tanks are very well-protected vehicles. Tanks have been destroyed by insurgents, but it has been a rare event that required very special preparation on their part to achieve. Some improvised weapons and man-portable rockets are effective against tanks, especially if fired at the sides or rear, but the problem with all of them is getting close enough to use them.
So, that calls into question just how cost-effective doing something like this would be. Is that really a good investment of defense dollars? Not if it's going to create all kinds of problems in the process which - right now - it would. Therefore, this is something we can't expect to see any time soon. This is not meant sarcastically, but if we're still talking in 20 years, we might be seeing the infancy of this then.
Quote:
To those of you unfamiliar with the series, Dale Brown, an ex-fighter pilot, has been writing novels featuring repurposed fighters and bombers since 1987's "Flight of the Old Dog". I just finished the latest in the series. Rebuilt as drones fighter jets have featured in many if the novels.
Dale Brown's novels can be fun to read, but they all border on the absurd, in the technology, tactics, and plots. His entire universe runs on "rule of cool". Most every cool project he talks about would be facepalm-worthy expensive if actually tried, and I have no idea what an actual aeronautical engineer might think of the more outlandish stuff.
Login
Register
FAQ
Search
