/sci/ - Science & Math

[Anonymous No.7563974]

It's more than 30 years since the original Terminator. Why can't robotics build a functioning T800 yet? Maybe without the CPU, but something you can remotely control that can stand and walk on its own.

[Anonymous No.7563990]

>>7563974
You can. The walking part is hard though.

[Anonymous No.7563994]

>>7563974
We're getting there...

https://www.youtube.com/watch?v=mi0sX2Dxk6o

https://www.youtube.com/watch?v=knRyDcnUc4U

I'd say 20 more years.

[Anonymous No.7563995]

>>7563974
I think people suck at this.
I don't see what would be so hard about creating one. You would need a lot of sensors. Balance sensors, and you would need muscles that work smoothly and fast. Then you need to include all the mass distributions of the body arms and legs to get accurate responses to sensory input.

[Anonymous No.7563996]

>>7563974
https://www.youtube.com/watch?v=cNZPRsrwumQ

it's been around for at least 5 years now

[Anonymous No.7564018]

>>7563995
Each of which is a hard engineering problem, let alone coordinating and assembling each part.

[Anonymous No.7564020]

>>7564018
Only if you're a dumbass tbh

[Anonymous No.7564031]

>>7563974
we still have until 2029.

[Anonymous No.7564038]

>>7563995
>I don't see what would be so hard about creating one
Because you know little about the problem domain?

[Anonymous No.7564065]

>not knowing that Japan has been creating an army of t800's under the streets of Tokyo for the past 20 years all in hopes of recreating mechs-hitler and begin the next robo-reich

Am I reich or am I reich...

[Anonymous No.7564068]

>>7564020
>>7563995
typical teenager thinking he can do a better job than people working on these problems for their entire careers

[Anonymous No.7564071]

>>7564068
Not a teenager, and yes that is often the case.

[Anonymous No.7564093]

>>7564071
okay kid

all i have to say to this is "then do it"

[Anonymous No.7564097]

>>7564093
As I said, I'm not a kid, retard.
I won't do something simply to prove that I can, wasting years of my life. I have other things to do.
That being said, I probably wouldn't mind working on this. But I wouldn't be willing to spend the money required.

[Anonymous No.7564108]

>>7563974
simple as that: there's enough brain matter for the job, but not enough budget. no one is paying much enough to accelerate the process. and this is because human-like robots don't have much economical usage. industrial bots are designed to do their jobs and they do it better than any human (or human-like) could. so why should anybody invest?

[Anonymous No.7564113]

>>7564097
then shut up

[Anonymous No.7564178]

>>7564113
Aw, are you a little butthurt? Shut up yourself, little powerless bitch.

[Anonymous No.7564180]

>>7564178

[Anonymous No.7564197]

>>7563974
The basic answer comes in two parts: The computational difficulties, and the mechanical ones.

Computationally speaking, walking is actually incredibly difficult. It only seems easy because evolution has seen fit to equip you with an extremely large massively parallel neural net processor adapted through millions of years of evolution to do precisely that. Inputs from many sensors must be analyzed and correlated, and based on that, and this must be used to control a kind of barely-controllable stable falling in complex and changing conditions on a millisecond-to-millisecond basis without falling over. The brain and nervous system are terribly efficient at this, especially in terms of power consumption - the entire brain consumes just 25 watts, about as much as a low-end laptop doing nothing particularly intensive. You want to do it in a robot, untethered? You'll need to bring a pretty high-powered processor, cooling system for that processor (heavy), and add a hefty bit of mass to your power supply to handle both of those things.

Mechanically speaking, robots are currently designed almost nothing like humans. Human walking is optimized for the human body, which is a very different thing from robots as we can currently design them. The human gait is still incompletely understood, but appears to be a rather fluid and elegant process involving elastic energy and chanelling momentum and gravity to do your work for you, managed through a design extremely different from the rigid, rotating-joint devices we can build. The actuators are very different, too - motors have very different (and in some cases, worse) characteristics than muscles, and pneumatic/hydraulic actuators are closer but have their own disadvantages. And they are all quite heavy, which means the frame needs to be heavier and stronger, and which means the power supply needs to be even bigger to handle it, which means yet more weight.

[Anonymous No.7564211]

>>7563995
>I don't see what would be so hard about creating one
>meanwhile the MIT leg lab has been working on this since the 80s
https://www.youtube.com/watch?v=XFXj81mvInc

fucking high school dropouts and their delusions of grandeur
yeah let's just shove a bunch of sensors and actuators into a chassis and let it walk, it's that easy :^)

[Anonymous No.7564225]

>>7564211
See, that is better. You gotta agree that some of the other robots you see are shit.
Take this thing a step further and you've done it. I told you it wasn't hard.

[Anonymous No.7564240]

>>7563974
It's been 120 years since H.G. Wells' The Time Machine

Why can't we travel to a ruined ultrafuture and save slaved humans from their tunnel crawling masters?
ITS 2015 COME ON
TWO THOUSAND
FIFTEEN

[Anonymous No.7564247]

>>7564240
H.G. Wells is obviously a retard who knows nothing about science since he thinks time-travel is science-fiction and not pure impossible autism.
Robots on the other hand existed since the discovery of electricity and has been developing since.

[Anonymous No.7564264]

>thinking that building something capable of simulate the basics of human function is that easy

Someday they will do it.

[Anonymous No.7564302]

>>7564225
you are a top tier retard

>watches video
>"well it took me a few minutes to watch so it must have taken a few minutes to make! that was easy...."

[Anonymous No.7564307]

TARS would probably be more feasible

[Anonymous No.7564334]

>>7564302
You are seriously top tier retard if that's your conclusion of what happened.

[Anonymous No.7564702]

>>7563974
ATLAS, not terminator.

[Anonymous No.7564709]

>>7564197
You are right, but for flat terrain walking is a largely solved problem:
http://www.youtube.com/watch?v=YvbAqw0sk6M
>>7564211
I think it's best to tease the teens to get into hobby robotics than bash them. 1 in 10 may even do a real robot like this: http://www.youtube.com/watch?v=JNzUC5VBF8Y

[Anonymous No.7564723]

>>7564240

[Anonymous No.7564730]

>>7563995
That's it? He'll I'll get started right now after a few games of Halo

[Anonymous No.7564731]

>>7564178
Damm dude. You sure told him. Now he looks like a fool

[Anonymous No.7565311]

>>7564225
>I told you it wasn't hard
oh for fucks sake

[Anonymous No.7565330]

>but something you can remotely control that can stand and walk on its own

Design the remote control.

[Anonymous No.7565376]

>>7564709
Oh, sorry. I'd just assumed he meant "walk like a human" in terms of adaptability to terrain, etc. An earlier revision of that post mentioned that we had actually pretty much solved flat walkers as long as you don't also need them to run, but I hit the character limit and cut it.

[Anonymous No.7565395]

>>7563974
>something you can remotely control that can stand and walk on its own.
The army has been doing this for the last decade or so.