NationStates

Dtn wrote:if a mecha doesn't have a sword or axe it should at least have 16" guns

Like this? sword √ 16" guns √ mecha √

Select your layout.

New Vihenia wrote:Select your layout.

(Image)

The first one is more compact and logical, second is... Completely nonsensical and makes the tank a much bigger target from the front and back. Not to mention difficulties in turning and transporting it via rail, aircraft or even naval craft. Plus the sheer size means that even if you could carry it via those means (not referring to rail, even with a comically massive rail gauge) you'd be carrying far fewer tanks in that regard with more space taken up by these wide bricks.

top is cutest

bottom is too wide imo

Antipatros wrote:An idea that you're seeing with some future tank designs is increased drone integration (for example, a tank launching small quadcopter drones, or mixed platoons of unmanned and manned tanks).

Would it make sense to add a crew member who helps to manage/control these drones, or should the tank commander be tasked with this? Many countries used to have radio operators in their tanks (who also often controlled bow machine guns), but then removed that position later on.

I've ran into the idea of letting the driver play with that when not actually moving the tank.

But then it depends on what sort of drone it is. If it's just a glorified helicopter-mounted periscope with autopilot, then there's not really much added workload and it'll improve the driver's situational awareness and the TC's if the TC gets a video-feed of what the driver sees the same way they can see what the gunner sees.

That way the TC gets the situational-awareness without the added workload.

The downside is now you may end up with a driver trying to pilot 2 or 3 vehicles simultaneously whilst being shot at.

Although most UGVs have a "follow the leader"/Convoy AI system so they can be collectively controlled with some ease. And Quadcopters can be virtually tethered to fly within a certain distance of the battlegroup at all times.

Sometimes it's a moot point since the control-vehicle is often an APC following behind the battlegroup.

Hurtful Thoughts wrote:

Antipatros wrote:An idea that you're seeing with some future tank designs is increased drone integration (for example, a tank launching small quadcopter drones, or mixed platoons of unmanned and manned tanks).

Would it make sense to add a crew member who helps to manage/control these drones, or should the tank commander be tasked with this? Many countries used to have radio operators in their tanks (who also often controlled bow machine guns), but then removed that position later on.

I've ran into the idea of letting the driver play with that when not actually moving the tank.

But then it depends on what sort of drone it is. If it's just a glorified helicopter-mounted periscope with autopilot, then there's not really much-added workload and it'll improve the driver's situational awareness and the TC's if the TC gets a video feed of what the driver sees the same way they can see what the gunner sees.

That way the TC gets the situational awareness without the added workload.

The downside is now you may end up with a driver trying to pilot 2 or 3 vehicles simultaneously whilst being shot at.

Although most UGVs have a "follow the leader"/Convoy AI system so they can be collectively controlled with some ease. And Quadcopters can be virtually tethered to fly within a certain distance of the battlegroup at all times.

Sometimes it's a moot point since the control vehicle is often an APC following behind the battlegroup.

Adding to that, most localized UAVs like Fire Scout are readily deployed independent of the vehicles. In the case of Switchblade, a single man could carry it into the field and launch it themselves, though it's better to have two men work as a team to make things more efficient, again, independent of any vehicles in the area, freeing up space on those vehicles for things like APS or remote-controlled machinegun turrets... Or even electronic periscopes for the commander. Or a combination of these things.

Gonswanza wrote:

Hurtful Thoughts wrote:I've ran into the idea of letting the driver play with that when not actually moving the tank.

But then it depends on what sort of drone it is. If it's just a glorified helicopter-mounted periscope with autopilot, then there's not really much-added workload and it'll improve the driver's situational awareness and the TC's if the TC gets a video feed of what the driver sees the same way they can see what the gunner sees.

That way the TC gets the situational awareness without the added workload.

The downside is now you may end up with a driver trying to pilot 2 or 3 vehicles simultaneously whilst being shot at.

Although most UGVs have a "follow the leader"/Convoy AI system so they can be collectively controlled with some ease. And Quadcopters can be virtually tethered to fly within a certain distance of the battlegroup at all times.

Sometimes it's a moot point since the control vehicle is often an APC following behind the battlegroup.

Adding to that, most localized UAVs like Fire Scout are readily deployed independent of the vehicles. In the case of Switchblade, a single man could carry it into the field and launch it themselves, though it's better to have two men work as a team to make things more efficient, again, independent of any vehicles in the area, freeing up space on those vehicles for things like APS or remote-controlled machinegun turrets... Or even electronic periscopes for the commander. Or a combination of these things.

I'd still give the driver access to the TC's periscope/CITV for better obstacle-avoidance.

Worst case, give him remote-override of a CROWS-mount if they get immobilized and they need to lay down some lead.

Hurtful Thoughts wrote:

Gonswanza wrote:Adding to that, most localized UAVs like Fire Scout are readily deployed independent of the vehicles. In the case of Switchblade, a single man could carry it into the field and launch it themselves, though it's better to have two men work as a team to make things more efficient, again, independent of any vehicles in the area, freeing up space on those vehicles for things like APS or remote-controlled machinegun turrets... Or even electronic periscopes for the commander. Or a combination of these things.

I'd still give the driver access to the TC's periscope/CITV for better obstacle-avoidance.

Worst case, give him remote-override of a CROWS-mount if they get immobilized and they need to lay down some lead.

Okay, yea, fair enough. Plus all that would require, at most, is some extra wiring and some more spare parts (back @ the base) that allow for easier, quicker repairs.

This is of course better than ripping open a chunk of the hull to install a launch system for a UAV, plus a control setup and the bungle of cables that would come of that... Along with some mild complications that could arise.

Hurtful Thoughts wrote:I'd still give the driver access to the TC's periscope/CITV for better obstacle-avoidance.

Worst case, give him remote-override of a CROWS-mount if they get immobilized and they need to lay down some lead.

Just give the driver an MFD with access to the gun camera like in the CV90 tbh.

Austria-Bohemia-Hungary wrote:

Hurtful Thoughts wrote:I'd still give the driver access to the TC's periscope/CITV for better obstacle-avoidance.

Worst case, give him remote-override of a CROWS-mount if they get immobilized and they need to lay down some lead.

Just give the driver an MFD with access to the gun camera like in the CV90 tbh.

No, I want the driver to get his bow machinegun back... except it's mounted on top of the turret, and gyro-stabalized... and comes with a thermal sight.

If he gets disoriented he just needs to look out the hull viewport.

Re the "tanks with drones" idea: If your tanks already have a system like the RL American 'IVIS' (see, for example, https://www.military.com/equipment/m1a2 ... attle-tank) then you could just make data from separately-controlled drones -- perhaps operated from back-up AFVs belonging to the same units -- accessible through that...

Gonswanza wrote:Adding to that, most localized UAVs like Fire Scout are readily deployed independent of the vehicles.

This is unsurprising since Fire Scout is a brigade-level asset.

Hurtful Thoughts wrote:I'd still give the driver access to the TC's periscope/CITV for better obstacle-avoidance.

I've often wondered about the absolute worst way to make use of all the doohickeys on tanks myself.

Gonswanza wrote:This is of course better than ripping open a chunk of the hull to install a launch system for a UAV, plus a control setup and the bungle of cables that would come of that... Along with some mild complications that could arise.

Not sure about the "bungle of cables," modern armored vehicles already have computers and touchscreens.

People are over-complicating this question by over-simplifying it. Integration of unmanned systems with tanks covers a wide range of systems that will be controlled in different ways. There's no need for a tank to launch and control a 20-foot helicopter or Predator drone, nor is there a need for an extra vehicle to launch a disposable camera to give a platoon commander a peek around a corner or over a hill. In the intermediate range it may be optimal for a tank crew to temporarily assume local control of company-level assets like UGVs or small UAVs.

The mysterious "Kyiv" is correct that most of this will be handled autonomously instead of by an extra crewman, leaving the existing crew to poke waypoints onto a touchpad and make shoot/no-shoot decisions.

tl;dr

Modern existing tank like M1A1 can just have the loader do all the UAS/UGV stuff and give it an autoloader to handle the ammo. It's probably easier to give the loader a toughbook and a rack of radios than removing the whole seat or something.

OTOH giving drivers periscopes or cameras beyond what's directly in front of them (even this can be somewhat superfluous depending on the TC) is bizarre space alien logic. The short answer is that the TC knows where he wants his gun, the driver doesn't need to worry about it, and he just needs to follow orders. A tank is commanded by the TC like an extension of his body or a warship so the crewmen only need to know what is exactly pertinent to their own jobs.

Future tank like MGCS can have a three-man (or two-man?) crew with a TC looking through cameras of drones beaming pictures to him, or the gunner/TC operating a Black Knight-esque UGV ahead of them using their respective station controls, and driver maintains local SA. Maybe it beeps at them if it sees a dude with an RPG or something but that seems somewhat unreliable. In practice given things like Longbow's automatic target recognition algorithms routinely pinged random sand dunes and electrical towers in Iraq as SA-8s because the sort of calibration requires very extensive local terrain mapping and surveying to work right, so I'm just a bit skeptical of ATR in general.

This is probably what Dumbla does anyway.

Gallia- wrote:tl;dr

Modern existing tank like M1A1 can just have the loader do all the UAS/UGV stuff and give it an autoloader to handle the ammo. It's probably easier to give the loader a toughbook and a rack of radios than removing the whole seat or something.

OTOH giving drivers periscopes or cameras beyond what's directly in front of them (even this can be somewhat superfluous depending on the TC) is bizarre space alien logic. The short answer is that the TC knows where he wants his gun, the driver doesn't need to worry about it, and he just needs to follow orders. A tank is commanded by the TC like an extension of his body or a warship so the crewmen only need to know what is exactly pertinent to their own jobs.

Future tank like MGCS can have a three-man (or two-man?) crew with a TC looking through cameras of drones beaming pictures to him, or the gunner/TC operating a Black Knight-esque UGV ahead of them using their respective station controls, and driver maintains local SA. Maybe it beeps at them if it sees a dude with an RPG or something but that seems somewhat unreliable. In practice given things like Longbow's automatic target recognition algorithms routinely pinged random sand dunes and electrical towers in Iraq as SA-8s because the sort of calibration requires very extensive local terrain mapping and surveying to work right, so I'm just a bit skeptical of ATR in general.

This is probably what Dumbla does anyway.

Well, I was picturing the TC just ordering the driver/drone-operator to move doohickey A over the ridge so we can see what not to do when moving doohickey B to said ridge.
(although realistically if you're using 3-person tanks you probably went with an APC loaded with drone-operators and the periscope is just a really tall CITV)

Or to pull security when the tank is immobilized. (mostly this)

One is to reduce overwork on a 2/3-person crew, the other is to maintain SA when the rest of the crew is busy.

The drone obviously drives or flies itself! It just needs someone who can look at what it's looking at semi-continuously, which I suppose the TC can do if it beeps at him when it sees something neat, but it would probably be better if you had another person do this job instead considering the TC has his own optical sights to look through and is telling the gunner where to look. That person can then tell the TC to look at this neat looking thing.

Also giving a driver a view higher up/further back is pretty much the exact opposite of what you want to do for "obstacle avoidance" lol.

Gallia- wrote:The drone obviously drives or flies itself! It just needs someone who can look at what it's looking at semi-continuously, which I suppose the TC can do if it beeps at him when it sees something neat, but it would probably be better if you had another person do this job instead considering the TC has his own optical sights to look through and is telling the gunner where to look. That person can then tell the TC to look at this neat looking thing.

Also giving a driver a view higher up/further back is pretty much the exact opposite of what you want to do for "obstacle avoidance" lol.

Depends on how well drones fly on their own. By same logic the tank could drive itself, after-all. Thus freeing the driver to just be a human-override when things get tricky.

Except when using it as a rearview mirror.

It isn't going to be very long before tanks are driving themselves.

Austrasien wrote:It isn't going to be very long before tanks are driving themselves.

Under the control of Skynet, of course...

Why use a tank when the requirement is firepower, not numbers?

A-10 CAS or LCM-8 gunships will inevitably be used.

This is a modern siege, but one will find that only advantage tanks have is operating cost per hour:

https://en.wikipedia.org/wiki/Battle_of_Khe_Sanh

Hurtful Thoughts wrote:By same logic the tank could drive itself, after-all.

Yes! It's almost as if FCS was the F-4 of AFVs to some nebulous concept's (i.e. Carmel's) F-35.

The only purpose of a drone or tank "crew" at the end of a day would be to observe its sensors, and tell it to shoot things, to ensure it isn't going to be caught with its pants down and make sure the human troops aren't walking into an ambush. Or more importantly, that it doesn't commit an embarrassing mistake that has to be covered up because the robot's ATR mistakes a camera for a RPG, because a "killer robot" will be even worse a PR disaster than "crazy Vietnam vets" or "baby killer gunship pilots" that could cost real people their actual jobs.

There is a similar thing happening in airplanes, except it's been here since the 1990's, because flying a plane is significantly easier and simpler a task than driving a truck in terms of automation (and it's probably that much harder to do manually).

Gallia- wrote:

Hurtful Thoughts wrote:By same logic the tank could drive itself, after-all.

Yes! It's almost as if FCS was the F-4 of AFVs to some nebulous concept's (i.e. Carmel's) F-35.

The only purpose of a drone or tank "crew" at the end of a day would be to observe its sensors, and tell it to shoot things, to ensure it isn't going to be caught with its pants down and make sure the human troops aren't walking into an ambush. Or more importantly, that it doesn't commit an embarrassing mistake that has to be covered up because the robot's ATR mistakes a camera for a RPG, because a "killer robot" will be even worse a PR disaster than "crazy Vietnam vets" or "baby killer gunship pilots" that could cost real people their actual jobs.

There is a similar thing happening in airplanes, except it's been here since the 1990's, because flying a plane is significantly easier and simpler a task than driving a truck in terms of automation (and it's probably that much harder to do manually).

And this is why I laugh at MT nations trying to use fully automated armies of straight-up AI tanks and aircraft that would end up destroying themselves if the IFF cocks up much to the amusement of an enemy force that would have man-in-the-loop systems.

I don't know what's more likely, black budget project cheaping out and not spending another million hours of work into machine learning and target differentiation, and thus using organic computers (someone paired a mouse brain with a flight simulator in 2004) trained from birth to fight in wars...

Or...

Predator drones keep getting smaller and more numerous, and the enemy does the same.

Danternoust wrote:I don't know what's more likely, black budget project cheaping out and not spending another million hours of work into machine learning and target differentiation, and thus using organic computers (someone paired a mouse brain with a flight simulator in 2004) trained from birth to fight in wars...

Or...

Predator drones keep getting smaller and more numerous, and the enemy does the same.

The latter, due to how cheap and disposable UAVs are becoming, along with massive jumps in tech as well as machine learning.

Gonswanza wrote:

Gallia- wrote:
Yes! It's almost as if FCS was the F-4 of AFVs to some nebulous concept's (i.e. Carmel's) F-35.

The only purpose of a drone or tank "crew" at the end of a day would be to observe its sensors, and tell it to shoot things, to ensure it isn't going to be caught with its pants down and make sure the human troops aren't walking into an ambush. Or more importantly, that it doesn't commit an embarrassing mistake that has to be covered up because the robot's ATR mistakes a camera for a RPG, because a "killer robot" will be even worse a PR disaster than "crazy Vietnam vets" or "baby killer gunship pilots" that could cost real people their actual jobs.

There is a similar thing happening in airplanes, except it's been here since the 1990's, because flying a plane is significantly easier and simpler a task than driving a truck in terms of automation (and it's probably that much harder to do manually).

And this is why I laugh at MT nations trying to use fully automated armies of straight-up AI tanks and aircraft that would end up destroying themselves if the IFF cocks up much to the amusement of an enemy force that would have man-in-the-loop systems.

IFF doesn't exist in the future because the SAM system can literally identify a jet fighter better than a RAF plane spotter from ten times as far away?

An ATR would probably be no more prone to ambiguous resolution than a stressed human, at worst. Which means it's more than likely going to attack random civilians in the combat zone.

SA-8 warnings in Longbow's ATR were broadly ignored by crews when they occurred in Iraq (besides the initial one or two spooks by a random missile launch warning) because the mmW was spotting random corner reflectors and terrain multipathing it was confusing for Geckos. It wasn't properly tuned to Iraqi terrains (it was running a set of algorithms derived from in-depth observations in the Mojave) and that sort of algorithm used requires substantial manpower in tabulating data. But Longbow was designed in the 1980s and its ATR algorithms are extremely primitive by modern standards. Modern Longbow TADS and radar detect objects by looking at more than just edge detection (which is also called contrast detection), and futuristic ATRs will probably be able to determine exactly what sort of object a person is holding based on hyperspectral signature of the polymers or metals used in the construction of the object.

That might mean cameramen will be attacked because they're identified as holding a man-portable air defense laser or something though but that seems unlikely since a laser and a camera are probably pretty different.

It's more likely that a human will tell a robot to vaporize a house or something that the robot sees civilians in because they think there's a sniper or something in there because the robot tank is telling them that gunfire is coming from roughly that direction, with an ambiguity of 50 meters or so. Wouldn't be the first time a fully robotic combat system was told to engage something it (correctly) didn't recognize as a threat.

But due to cultural concerns, it's unlikely people will be willing to fully relinquish combat to robotic domains, despite the potential for far greater accuracy in target identification. Yet.

Regardless, it falls back to the question of "should we even have an entirely automated army, navy and air force?" To which the reply is almost always "well, no, because even tech has it's limits and there will be problems in one way or another, multiplied by the sheer size of the forces present" or something to that degree.