StompingOnTanks, on 05 December 2013 - 06:39 PM, said:
I think what he meant was autonomous mechs, not aerial drones... Plus there are autonomous drone tanks in the works too.
Autonomous drones don't have the flexibility or creativity of the human mind. Sure, you can build automated supply vehicles that follow a pre-programmed course, once on that path, the AI may be smart enough to stop the convoy in the event of an obstruction in the route, and may be smart enough to plot an alternate route, but the moment that same convoy is ambushed, it's good as lost.
Automated weapon systems are even more difficult, considering most warfare is urban in nature, and such a system would be unable to distinguish between non-combatants and guerrillas.
Edited by Vanguard319, 07 December 2013 - 04:55 PM.
Vanguard319, on 07 December 2013 - 04:50 PM, said:
Autonomous drones don't have the flexibility or creativity of the human mind. Sure, you can build automated supply vehicles that follow a pre-programmed course, once on that path, the AI may be smart enough to stop the convoy in the event of an obstruction in the route, and may be smart enough to plot an alternate route, but the moment that same convoy is ambushed, it's good as lost.
Automated weapon systems are even more difficult, considering most warfare is urban in nature, and such a system would be unable to distinguish between non-combatants and guerrillas.
Very good points. I personally don't believe soldiers will ever be completely replaced on the battlefield. Manned fighter planes could become a lot less common though (and there go my dreams of being a fighter ace. T_T Oh well, I still have Ace Combat and an Xbox.)
The real attraction of mechs in the battlefield, is not to pilot legged vehicles but to replace lives with automated machines. Also, giant mechs and autonomous machines are not exclusive, but ultimately complimentary. There is going to be some controller in the battleflield that can control a squad of mechs, Mechcommander style. Or perhaps there is going to be a piloted mech, controlling a company of autonomous mechs, not unlike you in Mechwarrior 4 controlling your NPC wingmen in their mechs.
The real attraction of mechs in the battlefield, is not to pilot legged vehicles but to replace lives with automated machines. Also, giant mechs and autonomous machines are not exclusive, but ultimately complimentary. There is going to be some controller in the battleflield that can control a squad of mechs, Mechcommander style. Or perhaps there is going to be a piloted mech, controlling a company of autonomous mechs, not unlike you in Mechwarrior 4 controlling your NPC wingmen in their mechs.
That would seem very likely. If I were a mech pilot I certainly wouldn't mind having a squad of AI buddies with perfect accuracy and lightning-fast reflexes with me. The problem is anything that receives orders via electronic connection could possibly be hacked and turned against you.
StompingOnTanks, on 02 December 2013 - 06:40 AM, said:
This video is old, but that thing moves at a pretty decent pace. This was back in the mid 80's. Makes me wonder what might be stomping around in Area 51 right now...
That one actually went into production for logging machines to be used with difficult terrain/soft flooring;
I still Think the company Caterpillar could make a sweet functioning battlemech if it wanted to.
They certainly could try, with all the large-scale hydraulic equipment and such they've built. I doubt it'd be effective within the next 50 years though.
Mechs would also have a psychological advantage over tanks. Fighting a tank is scary enough, but it would be downright terrifying to have a 15 foot tall walking death machine stomping toward you. Also, I don't know if any of you are familiar with Heavy Gear or the Kuratas, but an interesting concept would be to have wheels (or treads) on the feet of the mech. This would allow it to have about the same mobility as a tank over flat terrain,so they could be deployed together if they needed to be. Another thing to think about is whether mechs would actually only have a single pilot or if they would have a pilot and a gunner like in modern attack helicopters or even as many as a tank which usually has a crew of 4. A mech would be pretty complicated for a single person to pilot (unless they use some kind of Brain-Computer Interface), so a crew of at least 2 would make more sense.
LocationTVM-Iceless Fold Space Observatory Entertaining cats...
Posted 05 April 2014 - 09:03 AM
Rushin Roulette, on 30 January 2014 - 06:35 AM, said:
That one actually went into production for logging machines to be used with difficult terrain/soft flooring;
Thats-a-what I say'ed
StompingOnTanks, on 02 February 2014 - 06:49 PM, said:
They certainly could try, with all the large-scale hydraulic equipment and such they've built. I doubt it'd be effective within the next 50 years though.
Some systems my be hydraulic assist, but servos, ring servos and artificial muscles are going to be the real work horses in walking tanks.
AWOL 01, on 04 April 2014 - 12:28 PM, said:
Mechs would also have a psychological advantage over tanks. Fighting a tank is scary enough, but it would be downright terrifying to have a 15 foot tall walking death machine stomping toward you. Also, I don't know if any of you are familiar with Heavy Gear or the Kuratas, but an interesting concept would be to have wheels (or treads) on the feet of the mech. This would allow it to have about the same mobility as a tank over flat terrain,so they could be deployed together if they needed to be. Another thing to think about is whether mechs would actually only have a single pilot or if they would have a pilot and a gunner like in modern attack helicopters or even as many as a tank which usually has a crew of 4. A mech would be pretty complicated for a single person to pilot (unless they use some kind of Brain-Computer Interface), so a crew of at least 2 would make more sense.
Wheels and treads in the feet, you have to remember the weight of a tank is much more outwardly distributed, that is a lot of ground crushing pressure on any type road surface. Walking Tanks would likely need be shipped on their backs or air dropped to combat locations. As for piloting, dont forget the amount of computer power that goes on between pilots and land/air/sea/space craft. Take some of the more modern fighter jets.. turn off fly by wire and its a deathtrap with too much power and too finicky aerodynamics.. so many micro adjustments too fine even for human comprehension at said speeds across so many control surfaces. Its not got to be able to read your brain (though we are making strides in that direction), its got to be able to read your body(we are also making great strides in this direction). What we need not is more efficient accesses to the brains inputs, our built in sensors have done us the service of getting this far (dont get me wrong I love all of my sensors) but, we will need to sometimes better understand them and to bypass them for speed of interface comprehension.
With the exception of only a few current Tech devices, the 4 or more legged walkers, are all too slow to be of any effect on a modern Battlefield. Speed and mobility are quickly becoming the only way to keep the unit alive to fulfill it's mission.
If the latest materials being tested for "muscle" replacement can actually be used in a larger scale, then we may see systems currently relying on hydraulics, pneumatics, electric motors all take a back seat to these more advanced elements. These new elements provide much finer control and in most cases out do any current systems in strength, dexterity and weight savings.
Having the ability to actually move these large constructions is all fine and dandy, but the key to having a functional machine is in it's sensor systems. The ability to "know" where it's parts are and what it needs to do in relation to it surroundings. Sensor systems that read the terrain around it and can apply the correct response to a chosen movement in whatever direction. Control over all the internal systems for movement, weight distribution, balance, and stability being only a few items.
The commercial devices that the public currently knows about are rather primitive in there ability to sense what is happening around them and how or if it should be responded to. Some of the DARPA machines are getting better with there ability to follow terrain and avoid obstacles. But they still lack the systems that can understand if the location they are about to place a leg or foot on is stable or capable of supporting the units temporary weight. In my opinion, this is where the advanced sensors and intelligence needs to be directed, understanding and knowledge of the environment around the unit. Currently most of these systems react "during or after the fact" when something fails in it's stability or balance. There are very few systems that can effectively "read and interpret" close terrain and understand what they are "seeing", in relation to what they are doing.
[Yes, the military does have terrain following devices for aerial units, with avoidance built in. But none of these units are actually required to "step on" that terrain.]
Side note:
The current LS3 or new "Big-Dog" chassis from the DARPA project and Harvard Dynamics, does show promise for what it's being designed for. A follow and load carrying unit. But still lacks the sensor smarts or intelligence to independently react to changing circumstances. A whole different level of "brain power" is required to know what to do and when. The 4 leg model works well for the role it is designed for and the sower speeds it operates at. Power requirements have changed it size a few times during it's testing and now the chassis is a rather large beast, about cow size now. If these new "replacement Muscles" could be used on that unit it may indeed enhance it functionality. (Google purchased Boston Dynamics, as one of at least 9 robotics related companies in the last year.!!)
Wheels and treads in the feet, you have to remember the weight of a tank is much more outwardly distributed, that is a lot of ground crushing pressure on any type road surface. Walking Tanks would likely need be shipped on their backs or air dropped to combat locations. As for piloting, dont forget the amount of computer power that goes on between pilots and land/air/sea/space craft. Take some of the more modern fighter jets.. turn off fly by wire and its a deathtrap with too much power and too finicky aerodynamics.. so many micro adjustments too fine even for human comprehension at said speeds across so many control surfaces. Its not got to be able to read your brain (though we are making strides in that direction), its got to be able to read your body(we are also making great strides in this direction). What we need not is more efficient accesses to the brains inputs, our built in sensors have done us the service of getting this far (dont get me wrong I love all of my sensors) but, we will need to sometimes better understand them and to bypass them for speed of interface comprehension.
I see what you're saying about the weight distribution, but a rolling movement would actually exert less pressure on the ground than walking would. For instance, when a human is running, the heel hitting the ground experiences about 3 times the pressure it does when walking. The mech would also be able to crouch while rolling to have a lower center of gravity and give it more balance. Or, if it's a 4-legged mech it would already have a very sturdy base. The wheel system wouldn't need to be anything fancy, just enough to get it to the battlefield where it would switch to a walking mode. The Brain-Computer Interface was more of just an idea. At the moment it really wouldn't be feasible, so it would make more sense to have a driver controlling the movement, a gunner, a navigator, etc. If you look at Boston Dynamics' Big Dog or Atlas robots, they are able to balance or catch themselves when pushed. Most of the mech's small adjustments when walking would be done by the computer, the same way they have to use fly-by-wire in modern fighters, like you said. The pilot would mostly be in charge of the basic commands that would then be supplemented by the computer's adjustments.
9erRed, on 05 April 2014 - 03:45 PM, said:
(Google purchased Boston Dynamics, as one of at least 9 robotics related companies in the last year.!!)
They're also doing a lot of research with AI. It's a good thing Google's on our side or we might have to start preparing for a battle with Skynet if you know what I'm saying.
LocationTVM-Iceless Fold Space Observatory Entertaining cats...
Posted 06 April 2014 - 08:49 AM
I wonder how Googles "Do No Evil" policy will hold true.. its been said BigDogs got one hell of a throwing arm now.. its going to be difficult to keep those projectiles 100% evil free.
At onetime it was a Company a day.
With the largest public listed purchase at 12.5 Billion $.
The list seems to start with online apps, then transition to devices, merge in sensor devices and software and outright robotic devices. [seems like remote driven, and independent movement, smart, robotic devices. Could we be looking at future earth bound or off planet devices?]
All of the brains within these company purchases, or the actual creators of whatever tech are signed into contracts that state they must stay with the companies for at least a year after Google quires them. So Google is collecting quite a few very smart players in the top end tech fields, and directing them for Google's endeavors.
2
StompingOnTanks, on 05 December 2013 - 06:39 PM, said:
