Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[alway]

I think comparing it to a tank is a false equivalence given that this is Japan. It would more likely be designed for non-military purposes such as search & rescue in the aftermath of a disaster or other humanitarian crisis. Though, again, military or civilian, most tasks which would justify the price tag can be accomplished better and cheaper by unmanned systems.

Here's the problem with mechsuits: For the cost involved, you are better off investing in a fully-mechanical system operated by remote control with AI assistance or near complete AI control. It minimizes the risk to humans, as these systems are far from invincible, while simultaneously cutting costs as it does not have to fit a person inside or protect said person.

What do you mean by fully mechanical system? A bipedal robot? I would think the costs for that would be far more than exoskeletons.

The purpose seems more for adding strength to ground troops. Troops that you cannot replace just because you have tanks and planes. We will always need a human involved right in the action simply because there are no robots for the foreseeable future that have the versatility of a human.

By unmanned system I do not mean 'a robot that looks like a man,' I mean 'a robot which was actually designed to do the task.' Or more likely, a large number of smaller robots. $750k can buy a lot of those. For recon within broken terrain, small robots capable of fitting through tight spaces and such will be king; for recon within intact terrain, flying robots like DARPA's hummingbird would be more effective. For carrying large loads over distances, something akin to Boston Dynamic's Big Dog would be more effective (and the US military plans to use the LS3 successor to it). Lifting materials is a bit more tricky simply because of the range of materials/weights, but again, can be done for much less than $750k/unit. Adaptive swarm robotic systems for moving materials are pretty much at the same point in research as exoskeleton systems; be it ground based or similar to quadrotors.

Similarly, I think you vastly overestimate the difficulty of AI systems for relatively narrow (as compared with deep problems like a strong AI, which are in fact very hard) problems like those associated with such tasks. Keep in mind, the modern approach to AI has been around for only a couple decades thus far; and it's already at the point where it's driving cars autonomously; a problem which is arguably as complex as those any of the aforementioned robots would be involved in. Especially since they would be probably only semi-autonomous, with high-level tasks planned by operators who are in the safety of somewhere which isn't a warzone or disaster area.

[rutsber]

I think comparing it to a tank is a false equivalence given that this is Japan. It would more likely be designed for non-military purposes such as search & rescue in the aftermath of a disaster or other humanitarian crisis. Though, again, military or civilian, most tasks which would justify the price tag can be accomplished better and cheaper by unmanned systems.

Here's the problem with mechsuits: For the cost involved, you are better off investing in a fully-mechanical system operated by remote control with AI assistance or near complete AI control. It minimizes the risk to humans, as these systems are far from invincible, while simultaneously cutting costs as it does not have to fit a person inside or protect said person.

What do you mean by fully mechanical system? A bipedal robot? I would think the costs for that would be far more than exoskeletons.

The purpose seems more for adding strength to ground troops. Troops that you cannot replace just because you have tanks and planes. We will always need a human involved right in the action simply because there are no robots for the foreseeable future that have the versatility of a human.

By unmanned system I do not mean 'a robot that looks like a man,' I mean 'a robot which was actually designed to do the task.' Or more likely, a large number of smaller robots. $750k can buy a lot of those. For recon within broken terrain, small robots capable of fitting through tight spaces and such will be king; for recon within intact terrain, flying robots like DARPA's hummingbird would be more effective. For carrying large loads over distances, something akin to Boston Dynamic's Big Dog would be more effective (and the US military plans to use the LS3 successor to it). Lifting materials is a bit more tricky simply because of the range of materials/weights, but again, can be done for much less than $750k/unit. Adaptive swarm robotic systems for moving materials are pretty much at the same point in research as exoskeleton systems; be it ground based or similar to quadrotors.

Similarly, I think you vastly overestimate the difficulty of AI systems for relatively narrow (as compared with deep problems like a strong AI, which are in fact very hard) problems like those associated with such tasks. Keep in mind, the modern approach to AI has been around for only a couple decades thus far; and it's already at the point where it's driving cars autonomously; a problem which is arguably as complex as those any of the aforementioned robots would be involved in. Especially since they would be probably only semi-autonomous, with high-level tasks planned by operators who are in the safety of somewhere which isn't a warzone or disaster area.

Those are a lot of good points. One thing with people in a room though, it might be hard to react to some things if these robots are in a warzone, such as a bomb going off. You wouldn't really be able to rely on reflexes for something like that. Also, if this is a wireless connection, you have the problem of a lost signal at a critical moment, or something akin to lag.

[Zangi]

Signal jamming technology... whenever someone develops a jammer for the signals used by unmanned drones and what-not... their feasibility drastically lowers for military applications. 
As for AI drones/mecha/bots... well, for now, you have a lot of complex stuff you would have to program for on the ground decisions that a normal foot soldier/field commander would have to make... but.. if all you are using it for is lifting stuff, yea that is fine.  Any other operations, like security or higher, I'd always have a trained human coupled with it, to make sure it is doing things right and would not be tampered with/tricked/false positives.

[Virex]

Also, the same tech that goes into power armor goes into smaller robots (artificial muscles, legged movement, portable power sources). It isn't completely an or/or situation.

[da_nang]

Obviously they need to hire Dr. Halsey.  From the future.

[Nadaka]

The day powered, bipedal armor becomes effective, I shall eat a tanker bar.

If we are talking about "GIANT ROBOT" power armor? I absolutely agree with you.

But the idea of a guy walking around with a full body small arms proof armor and the augmented strength and endurance to carry and use a light machine gun in the same way an unaugmented soldier uses an assault rifle? That is a far more interesting idea.

[kaijyuu]

Unmanned walking drones will happen long before powered armor.

[Armok]

I'm still waiting for someone to put a saddle and a machine gun on a BigDog

[Hubris Incalculable]

I'm still waiting for someone to put a saddle and a machine gun on a BigDog

Yes. This is obviously the best course of action.

[Nadaka]

I'm still waiting for someone to put a saddle and a machine gun on a BigDog

Here, take my sword.

http://www.popularmechanics.com/technology/gadgets/4258963

[Armok]

That thing is neat, but I meant a legged and/or manned one.

[MetalSlimeHunt]

By the way, just so we're clear, the Liberal Democratic Party is the major conservative faction and the Democratic Party is the major liberal faction. Yes, very confusing.

[Starver]

Until then, whatever a walker can do a tank can do better.

Your comparison reminds me of this...

[Itnetlolor]

They have already worked on stuff like this in the private sector for a while. Sorta incredibly cheesy, but it's a commercial product not a military demonstration. Would not be too much of a stretch to adapt it to more durable materials. Probably won't meet military needs without the power generation thing the US currently struggles with.

Why does this technology remind me of Aperture Science? Must be the aesthetics.

I also find it funny that they're called Cyberdyne Inc.. I'll be wary of their terminators when they roll walk about.


Where battlesuits are concerned, why aren't they working on an Iron Man model? Somebody has to be enough of a fan to work on one for the hell of it.

[RedKing]

Here's the thing nobody is talking about, that jumped out to me:

In 2008, the Japan Science and Technology Agency estimated that the cost to build a fully function Gundam-like robot would be $725,000 USD for parts and material. Labor and other costs were not factored into the estimate.

What. I'm not sure the Pentagon can build a goddamn porta-potty that cheap. Certainly not an MBT. For the unit cost of a single Abrams tank, you could build 11 of these "Gundams", if that figure was accurate.

There's also logistics and scaling issues that could make this a somewhat rational idea to pursue. The recent engagement in Iraq, for instance, showed the drawbacks of using APVs and MBTs for patrol and garrison duty. This would seem to occupy a niche between standard infantry and personnel carriers. You get the armor and heavier weaponry of an APV, but disperses the potential targets so that if someone uses an IED or rocket or what have you, you maybe lose one guy intead of a whole wagonload.

Of course, I'm just absolutely geeked about the idea of a Japanese Battlemech Defense Force, so I may be biased. 
The real kicker is when they unveil the genetically-engineered army of blue-haired catgirl clones with psychic powers they've developed to pilot them.