[a / b / c / d / e / f / g / gif / h / hr / k / m / o / p / r / s / t / u / v / vg / vr / w / wg] [i / ic] [r9k / s4s / vip / qa] [cm / hm / lgbt / y] [3 / aco / adv / an / asp / bant / biz / cgl / ck / co / diy / fa / fit / gd / hc / his / int / jp / lit / mlp / mu / n / news / out / po / pol / qst / sci / soc / sp / tg / toy / trv / tv / vp / wsg / wsr / x] [Settings] [Home]
Board
Settings Home
/m/ - Mecha


Name
Spoiler?[]
Options
Comment
Verification
4chan Pass users can bypass this verification. [Learn More] [Login]
File[]
  • Please read the Rules and FAQ before posting.
  • There are 108 posters in this thread.

05/04/17New trial board added: /bant/ - International/Random
10/04/16New board for 4chan Pass users: /vip/ - Very Important Posts
06/20/16New 4chan Banner Contest with a chance to win a 4chan Pass! See the contest page for details.
[Hide] [Show All]


Janitor applications are now closed. Thank you to everyone who applied!



File: cockpit.jpg (80 KB, 640x355)
80 KB
80 KB JPG
What is the best way to control a mecha and why do you think it could realistically work?
>>
>>16139551

Probably something akin to the AV system in IBO, where your body becomes the brain of the giant robot and you directly control it's movements via thought rather than indirectly through pedals, levers or hand off control to an AI or macros.
>>
>>16139551
>What is the best way to control a mecha
With one lever and three buttons. When you push the lever forward, the robot goes wherever you want. When you push any of the buttons, the robot does anything else you want.

>why do you think it could realistically work?
I have performed exhaustive research of mecha control systems by watching thousands of hours of toy commercials, and this method works every time. There are no downsides.
>>
>>16139551
Motion capture for the leg, torso and basic arm-movements; eye-movement controlling where you aim at; and touchpanels attached to your wrists at a comfortable distance for your fingers to easily reach everything where you select weapons and stuff. Also a control board in front of you with switches and buttons for stuff you don't need while in combat.
>>
With a remote controll
>>
Keyboard and mouse, clearly.
>>
Most definitely the AV system or something similar that uses direct brain input. There is nothing better than 1:1 control.
>>
File: 20110710.jpg (80 KB, 1113x471)
80 KB
80 KB JPG
>>
File: g5MREr.gif (3.47 MB, 640x360)
3.47 MB
3.47 MB GIF
>>16139579
Pretty much this. A humanoid robot would be best controlled by the movements of a human body.
>>
>>16139563
/Thread
>>
'neural clamp' that prevents motor commands from being sent to the pilot's body and a VR type headset to feed the pilot video from the mech. then, using chemical magic, inject the pilot with some substance that increases reaction time/increases the pilot's perception of time to prevent the pilot's commanded movements and the mech's output from lagging or desynchronizing due to the size difference.
it makes sense if you don't think about it
>>
>>16139721
I'm fucking retarded and forgot the important part.
the clamp intercepts the motor commands and routes them into the mecha, and sensors on/in the hull send feedback to the pilot.
>>
File: 1508652678394.jpg (294 KB, 850x1200)
294 KB
294 KB JPG
>>16139551
>>
File: 1508652740998.jpg (202 KB, 850x1200)
202 KB
202 KB JPG
>>16139873
>>
File: 1508652803290.jpg (226 KB, 850x1200)
226 KB
226 KB JPG
>>16139875
>>
File: 1508652865814.jpg (203 KB, 850x1200)
203 KB
203 KB JPG
>>16139878
>>
>>16139551
You're all wrong. Two pedals, one for "forward motion" on either leg, like a tank. it allows you to move, turn, spin in place, and stop all without touching anything other than two pedals, one per foot. Heuristic AI machine learning computer science MIT labrat bullshit magicks away the need to actually control the legs in any more detail. Maybe a control panel can manually switch modes for shallow water, mud, cliffsides etc but walking should be automatic.

For arms, whatever is used to aim a gun will work. Whether a two handed contraption or two separste sticks, one per arm, just use whichever's best. If the thing has to do a job outside shooting, there can be a mission specific module that lets you do some kind of upper body mocap, like in Avatar, which may be useful in shit like rescue recovery and gassing colonies.
>>
I'd go with the cockpit setup from CCA, those spheres used to control the ms. Odd enough from something so hightech and streamlined, goes back to crowded cockpits and joysticks/levers in f91 and victory. Sometimes MS tech seems to take a step backwards.
>>
>>16139725
I figured that's what you meant. We're pretty much closing-in on that technology for prosthetic limbs: https://www.youtube.com/watch?v=wxIgdOlT2cY
>>
>>16139950
I guess it depends on the need for these features relative to cost. MS warfare became much smaller in scale after the Zeonic conflicts and the need to one-up the enemy died-down.
>>
File: poster_50.jpg (2.53 MB, 3465x4952)
2.53 MB
2.53 MB JPG
>>
File: white_20glint4_large.jpg (23 KB, 463x480)
23 KB
23 KB JPG
Pilot is linked neurally to mech for 1:1 control. Some method of paralysis is used (Possibly some gas added to their oxygen in life support) so that they don't move around in the cockpit. G-suit and other restraints are mandatory. Pilot has a sense of touch - and yes, that includes pain - so they have a better understanding of their environment.

Other than that I'm not 100% sure. I can already guarantee you that being paralyzed, restrained, and prone to feeling pain in a mech is someone's fetish, though.
>>
>>16139944
I'm an idiot, I don't mean "forward" i mean "forward backward". The pedals are not like car pedals, they're the ones you see in zeta, where the fulcrum is in the middle, more like a bike pedal. If you press with the toes/ball of the foot, it goes forward. If you press with the heel, it goes reverse.
>>
>>16139987
that's pretty much my idea but phrased better, neat
>>
>>16139551
Neurohelmet, duh.
>>
>>16139551
This is literally all you need.
>>
File: ms-06fz-cockpit[1].jpg (19 KB, 300x201)
19 KB
19 KB JPG
>>16139551
>>
>>16140055
This. Brain-Computer link is GOAT.
>>
exo sqad did it better, NEURAL JACK!
>>
File: Img174.jpg (442 KB, 2000x1330)
442 KB
442 KB JPG
>>16139551
like in Guilty Crown

Mechas having a control room
>>
Emotions, of course.
>>
>>16139987
Depending on how fast the mech moves and changes direction, it would make sense to ensure the pilot is completely restrained with mechanically controlled blood-flow. I envision a pod with a gel-padded body mold that constricts the limbs when needed to prevent blood pooling.
>>
>>16140297
That's a good if you're not a rapist who's angry at your friend.
>>
File: 48a.jpg (188 KB, 1600x1200)
188 KB
188 KB JPG
>>16139950
hey I love the CCA cockpits too, all the tech in that movie is so slick honestly
>>
Didn't Arm Slaves use some setup that combined motion tracing with controllers and buttons?
To me that sounds like the best you'd get, since then you'd end up with both the range of motion from the robot just moving how you move and the precision of a controller input.
>>
>>16139873
>that art

There's incest or horse fucking in this, isn't there?
>>
File: f-22 cockpit.jpg (84 KB, 735x931)
84 KB
84 KB JPG
>>16139551
Realistically, it'll probably be set up like your typical fighter jet cockpit like pic related, but most likely with an AI input of some kind to help compensate for movement.
>>
>>16139551

I always assumed Mobile suits had a lot of pre programmed motions that are activated with sticks like walking aiming and firing. anything that looks like a robotic like motion but it seems you can also use the sticks and the foot pedals to operate the limbs manually outside of these preset movements. the ability to make use of this on the fly would make the difference between an Ace and a grunt as the Aces always seem to perform moves that you wouldn't normally see anyone else do. stuff like the char kick and those kinds of motions.

generally it's accepted at least in Universal Century that mobile suits aren't terribly difficult to operate on a basic level considering how many civilian kids happen to fall into one.
>>
>>16140496
Only if you try to look further into it.
>>
File: traceon.gif (1.97 MB, 320x240)
1.97 MB
1.97 MB GIF
>>16139551
>>
File: metal_gear_op.jpg (18 KB, 559x314)
18 KB
18 KB JPG
>>16140047
>>
>>16139721
The problem with these kind of systems is how do you operate anything other than basic movement? How do you configure your radar, change radio channel, zoom in the camera, use your GPS maps etc. Most importantly, how do you turn it off?
Perhaps the best solution would be to have a second crewman to manage all the awkward stuff, like the gunner in a combat helicopter.
>>
>>16139551
https://www.youtube.com/watch?v=vUxDmKFCD2o
>>
>>16140552

How do you focus your eyes, tune out conversations you're not interested in or maintain balance? You do a lot of stuff automatically and without conscious thought through training as a child. The same would presumably be true of controlling a machine with your brain. The training isn't just to move it, but to control all its functions. Also, a safe gesture of some kind you would never normally make could be used to turn it off. Or just a thought, if thought is what normally controls it. Same as a person closing their eyes and going to sleep essentially.
>>
>>16139551
A combination of macros, motion controls, and thought recognition system. This is assuming it's a biped, if it's something like a tachicoma, macros, levers, buttons and voice commands might be enough.
>>
>>16140552
The problem is that any external plug or port interface will be prone to infection no matter what.
>>
Like a video game
Just stick an xbox controller in there and use the Armored Core FA setup
>>
File: gundam-300x214.jpg (23 KB, 300x214)
23 KB
23 KB JPG
>>What is the best way to control a mecha
The Senjou no Kizuna P,O,D control layout.

>>do you think it could realistically work
Yes, because it's basically a MS simulator that's been out in arcades and has managed to survive for over a decade. And to be honest, part of me truly believes its the Japanese government's way of training their population for MS warfare.

And all it has are two levers and two pedals. All you need are some preset command buttons (such as kneeling down) and you're good to go.
>>
>>16140678
Japan will never have a strong enough economy for mechs.
>>
>>16140683
They'll fund it with the money they earn from people playing the games! It's been over 10 years and its STILL 300 yen a game!Though I think it started off at 500 yen a game...
>>
File: Tachi.jpg (79 KB, 640x443)
79 KB
79 KB JPG
Depends on what you define as a mech here. Anything multi legged like pic related or MG Excelsus coul probably be controlled just fine with a conventional cockpit with pedals and maybe joysticks to control the "arms". Anything else is the AI's work.
If we're talking something bidep like the ST-84 or even Votoms where the arms are kinda crude yet important, maybe have direct commands from the pilot's arms, though joysticks work still.
If we're talking full retard Gundam where the mechs say fuck you to humanity and surpasses it in mobility, it's straight up brain/computer interface + AI assistance or bust.

tl;dr joysticks and AI.
>>
File: BLR Hardsuit.jpg (73 KB, 480x451)
73 KB
73 KB JPG
>>16139599
>>16139579
There are like two arms jutting out of this torso, I thought it was always cool for a pilot to control a mech like that
>>
>>16139551
Someone post the keyboard in Armored Core 3 Intro
>>
File: GOAGAINSTOTHERPLAYER.jpg (210 KB, 600x800)
210 KB
210 KB JPG
Virtual-On's thing of remote controlled robots is the best and cheapest solution. It works realistically simply because lack of casualties and you don't necessarily require brain surgery to operate it.
>>
File: robotjox4.jpg (128 KB, 620x338)
128 KB
128 KB JPG
>54 replies
>No one mentions the cockpits from Robot Jox

It's arguably the most realistic design in anything /m/-related. Blending mo-cap with joysticks and controls, some of which are in the pilot suit itself. It works because it marries the best of both worlds into a proper realistic combat machine.
>>
File: Shurikenjin_Cockpit.jpg (491 KB, 1920x1090)
491 KB
491 KB JPG
A tatami room where you swing swords around.
>>
File: oh_ffs.png (55 KB, 322x303)
55 KB
55 KB PNG
>>16140884
Did you even click the link in >>16140564? Apparently not.
>>
>>16139950
Those sphere control sticks were a bad idea to begin with and it even shows in the unicorn novel. They basically retconned them by saying they were outdated devices of the past, that the sleeves MS were plentiful but outdated.
Every other suit after the CCA period went back to the analog or whatever they call that horizontal control sticks most gundams have
>>
katamari controls
>>
>>16140899
Probably the most effecient way so far.

Does anyone remember that butler in G gundam who piloted a fencing mobile suit with looked to be the configuration of a car? Steering wheel and all?
>>
>>16140913
Actually, you're not quite correct. Federation MS had those spheres, they're called arm raker controls. Under each fingertip is a control switch, in addition to the grip itself being a type of joystick.

Neo Zeon MS in CCA might have also had arm rakers with the individual fingertip switches, but not shaped as the stupid half-sphere like the Federation models. They had cup-like guards that NZ pilots inserted their hands into to grasp the control sticks.
>>
>>16140913
>>16140949
I get that, still feel like spheres that control the roll pitch and yaw with button macros for arms, legs, fingers etc. is a simpler design than straight up flightsticks.
>>
I always wondered how you would control your legs to adapt to the specific terrain, the head, the fingers, each limb, the amount of power and force you move each limb.......
>>
Gunbuster, despite being fully 'super" and not trying to be realistic, has the best system. Two pilots. One has motion control and maps the body's actions to the robot's movements. The other sits on a control panel and inputs other commands.
>>
File: Grimdark.jpg (1.08 MB, 1750x1313)
1.08 MB
1.08 MB JPG
>>16139551
You must use the GRIMDARK set.

Alternatively you tell your supra advanced on-board computer what you want to do and with its sapience and super intelligence it completely ignore you and do the actual job while you push joystick believing you have any reasons to be here other than keeping the human in the loop.
>>
>>16140240
>NEURAL JACK!

The moment it jammed in was sweet.
>>
>>16140294
That's basically BCI(although without puting anything in your head) connected with remote control
>>
>>16141323
You don't. The computer does it through an incredibly complicated software suite comprising LIDAR, RADAR, and optical image terrain mapping, and control of the limbs by a sophisticated inverse kinematics model. You operate it like a one-man tank. You're still going to have highly complicated controls however, twin sticks with multiple hat switches with both rotation and likely sliding bases, and aircraft-like pedals that also use foot tilt as well as push.
>>
>>16139950
>Odd enough from something so hightech and streamlined, goes back to crowded cockpits and joysticks/levers in f91 and victory.
Not so. The early UC MS only had room for panoramic cockpits because they were so large. Late UC suits are much smaller, obviating the possibility of physically bulky panoramic cockpits from being properly viable.
>>
File: RADcover.jpg (45 KB, 275x384)
45 KB
45 KB JPG
>>16139551
>the best way to control a mecha
A Dual-Shock 2.
>>
I'm more concerned as to why nobody wears a damn seatbelt after Gundam up until Victory with it's airbelt. Buckle up ffs guys.
>>
>>16140621

Putting aside that neurel helmets that picked up signals without the need for a port or socket would make it a moot issue I don't really see how that's a major problem. Cleaning them would be a nuisance, sure, but it's not like it's the kind of over-riding problem that would stop the military doing it on it's own. Especially in a war setting. Shit, the British built planes out of wood soaked in gasoline to help the covering stick in WWII (I think that was the reasoning) and the majority accepted it just because beating the Germans was more important than risking burns if and when your plane got shot down. Which they often did.
>>
>>16139551
Airbus 320 jetliner controls. :^)
>>
>>16141655
yes having it piloted by remote control in a safe plase is better
>>
>>16142607
>Putting aside that neurel helmets that picked up signals without the need for a port or socket would make it a moot issue I don't really see how that's a major problem
There's not enough of a signal coming off a human brain for it to be reliable for combat, especially if any ECM is used.

Also you do not want an important asset like a pilot hardwired to the internals of the machine. A strong enough electrical charge will kill them. We're talking major surgery as well, so there's a nonzero chance of fatalities just from the operation to install the port. That's really not the kind of PR a military wants.

Try to put some thought into it instead of just reciting memes.
>>
There is only one answer.

J-Zyme jelly vat.
>>
>>16142146
What if I wanted to pickup a car with 2 fingers or rip my arm off and beat a zaku with it?
>>
>>16143156
You switch hand modes with a hat switch and lock onto it.
>>
File: gaiking 4.png (1.58 MB, 1255x708)
1.58 MB
1.58 MB PNG
>>
File: gaiking 6.png (1.85 MB, 1258x706)
1.85 MB
1.85 MB PNG
>>16143209
>>
>>16139551
Playstation 4 controller
>>
>>16143215
Die!!
>>
I have never understood how cockpits work in mecha. Especially those horizontal joystick ones that also function as accelerators
>>
File: Zerosys-interface1.jpg (33 KB, 512x384)
33 KB
33 KB JPG
>>16143342
this is why the zero system is the best, converting combat in simple timed yes/no scenarios
>>
>>16143355
The Zero system is essentially Quick time Events?
>>
>>16143359
basically

Press A to blow everyone up
or
Press B to slice everyone up instead
>>
File: gelgoogcontrols.jpg (168 KB, 707x1183)
168 KB
168 KB JPG
Here's how they lay out controls on a Gelgoog...
>>
File: hsgcw12_1_1462502659.jpg (35 KB, 600x450)
35 KB
35 KB JPG
>>16139563
It's kind of funny, but that's basically the official control setup of the mechatro WeGoes. You lean the control stick forward and backward to move it and the two buttons work like the controls on a claw game to move the arms.
>>
How about the build fighters one?
>>
File: 1301104747799.png (729 KB, 720x1280)
729 KB
729 KB PNG
You need to pilot by pole dancing
>>
Full metal panic meets 360 degree vision sphere.
>>
File: Trant-dies.jpg (30 KB, 512x384)
30 KB
30 KB JPG
>>16143359
No wonder it's so stressful.
>>
>>16139551
>why do you think it could realistically work?
One thing to keep in mind is with mobile suit technology there was large degree of automation pre-programmed as technology and combat-data moved on, like you wouldn't have to dick around with arm motion controls to swing a beam sabre as you'd have some pre-programmed attack macros bound to triggers on your sticks and it could use defensive measures automatically.
>>
>>16144692
>macros
I have a mini-seizure every time you faggots use that word. You just select the beam saber and push the hat switch in the direction you want to swing.
>>
>>16143131

> There's not enough of a signal coming off a human brain for it to be reliable for combat, especially if any ECM is used.

That's a nonsense argument. The strength of the signal coming off the brain has no real bearing on how viable a neural helmet is, which is mostly down to how sensitive the equipment measuring it is. There is already first generation stuff being developed capable of detecting and interpreting brain signals and using them to control other things.

https://www.ted.com/talks/tan_le_a_headset_that_reads_your_brainwaves#t-618116
https://www.ted.com/talks/greg_gage_how_to_control_someone_else_s_arm_with_your_brain#t-340406

Both those links have relatively cheap and simple technology that can detect brain signals and use them to control external objects. They might not be viable military equipment now, but give them a couple of decades and it's entirely possible given the leaps technology often makes once it's been set.

> A strong enough electrical charge will kill them.

So what? How is that any different to risking death by bullet, explosion, impalement or any of the dozens of other ways soldiers routine risk life?

> We're talking major surgery as well

Says who? Surgery to implant something in to the body can be routine; even stuff implanted to control major bodily functions. Pacemakers for instance. The risk is mostly determined by how experienced the people doing the procedure are and how good the equipment is; not by where it's going or what it's doing.

> That's really not the kind of PR a military wants.

See the aforementioned example of a the British Mosquito, which was a known pilot hazard due to the construction materials and which no one was bothered about because it was a total war situation and people found the need to repel invasion more pressing than health and safety.

> Try to put some thought into it instead of just reciting memes

What part of the idea of neural interfaces is a meme?
>>
this was the pilot mechanism from Char's Zaku II but it seems sort of hard to use. I don't think they would make something as convoluted as this irl
>>
File: 1421314198320.webm (2.86 MB, 1008x720)
2.86 MB
2.86 MB WEBM
If there's a problem with direct brain/machine interface, it's not the feasibility it's more the feedback loop.
We will have a gap between gaining the ability to do something as soon as we conceptualize it in our head... and the ability for computer to tell when you are not serious or want it don't another way.

So long as we use physical trigger we have inertia giving us time to counter involuntary reflex, but with thoughts you need something like a systematic delay that you can only override in advance or under specific condition.

By the time we can do that we might have invented Artificial Intelligence already and the singularity that come with it.
>>
>>16146407

> We will have a gap between gaining the ability to do something as soon as we conceptualize it in our head... and the ability for computer to tell when you are not serious or want it don't another way.

No, we won't. Humans are able to distinguish thinking about doing something and doing it. You can think about operating a muscle without actually doing it so why wouldn't you be capable of considering or thinking about firing a gun mentally without doing it?
>>
>>16139579
>>16139599
Kinda reminds me of FMP's Arm Slave system.

But I do think that a MoCap-esque system would certainly validate a humanoid design.
>>
>>16146413
Because we've got dedicated parts of our brain for controlling movement for those parts of our body. We don't have one for artificial limbs so you need to put one together artificially.

Now, we do have biofeedback systems but that's more you adopting a certain mindset. The brain doesn't exactly reserve parts of its'self for expansion so anything you try to get a read from probably has a use already.
>>
>>16146437

It doesn't really, because the brain is amazingly adaptable and relatively recent discoveries (re: neuroplasticity) essentially prove that the human brain will reassign less used areas of itself in order to improve parts that are used more. So if a human was strapped in to a dog shaped unit or octopus shaped one or something they'd eventually learn to start utilizing it properly if they used it enough and their brain would start reallocating space for things it doesn't need to ensure the pilot can fully take advantage of the system. The training to do so might take a year or two of dedicated effort, especially for more esoteric stuff, but it will happen. Same as people adapt to fake limbs, or even magnetic implants and so on. A humanoid design would be quicker to adapt, but that ease of adaption would need to be weighed against the forms advantages and could come off the worse because training time might be acceptable for other forms if they grant enough advantages of their own.
>>
>>16146465

> The brain doesn't exactly reserve parts of its'self for expansion

That's exactly what it does do though. See the previous post and neuroplasticity. The brain doesn't come with any dedicated parts. That's why babies struggle and need to learn, because the brain is building a database through experience. You might need to get a bit of experience to differentiate just thinking about doing something in a giant robot to actually doing it, but you will be capable of controlling it; just like you do your own body. Your own body is basically just a suit to protect the brain that it learns to adopt to over the years as is.
>>
>>16146413
You miss the point completely,
You are not the one needing to know wether or not it's serious, the computer do and IT ISN'T US, it will not be reading "yes" or "no" it will need to search others clues/thought to know if what it read is 51% sure to be an order to shoot or 49% sure to be an order to shoot IF AND ONLY IF the target make a move.

Anything you can do to prevent error will run against your needs. You don't want a 2 second delay in a life an death situation.

Aside, it take extreme training to actually control your body. The difference is that instead of breaking a porcelain vase in an involuntary movement your mech will kill kill someone.


While I'm on the topic.
Take Pacific Rim control interface, the one that look stupidly contrived for some watcher.

Having two/three computers work in parallel and crosscheck their data is already a time approved method to prevent errors.
Modern Airliner have procedure where to override the computers imposed limitations the two pilot have to simultaneously pull the joystick in order to bring the true force of the control command.
Which bring me to this link:
http://journals.plos.org/plosone/article?id=10.1371/journal.pone.0020422
In short: if a computer read the mind of two persons trying to think of the same, it go from 60% of getting it right to 95%.
>>
>>16146547

> the computer do and IT ISN'T US, it will not be reading "yes" or "no"

Except anyone piloting one will have trained within the system to give it the information it needs to distinguish the brain pattern for "I'm thinking about doing this but don't intend to actually carry it out" and "I want to do this". Just like the guy using the system had in the TED Talk in >>16144750 (https://www.ted.com/talks/tan_le_a_headset_that_reads_your_brainwaves#t-618116) had to set up the system to recognize his brain patterns in certain situations.

> Aside, it take extreme training to actually control your body.

So? It takes years of training to become a pilot too, including lots of physical training to be able to withstand certain things. Just because it's extreme doesn't mean people are against it.
>>
>>16146570
Don't argue with the idiot. He's one of those people who doesn't accept limits.
>>
>>16146407
>singularity
Called it, he's an idiot.
>>
File: 1512636850011.jpg (302 KB, 1265x1563)
302 KB
302 KB JPG
>>16146570
>give it information
You want to somehow censor your own thoughts so you never think anything that could be misinterpreted by a computer and teach the system to recognize what it should ignore what and when?
That's precisely the point!
You can't use mindreading for control interface if it risk misunderstanding you and going haywire or stay stuck in indecision until it get a clear answer. The TED talk you linked to make my point.

>So? It takes years of training to become a pilot too, including lots of physical training to be able to withstand certain things. Just because it's extreme doesn't mean people are against it.
...and the primary factor in disaster is human error. ERROR, not incompetency.
The first method to deal with this is to replace pilot by automated system, the second is to force the user to go through numerous check and procedures.
The first is the opposite of what you want H/M interface for and the second is incompatible with the reaction time you want.

By the time you solve this problem it will be more efficient to have a robot do your job.

>>16146683
>>16146686
You can be easily replaced by a shitposting bot. How did you even get through the CAPTCHA?
>>
Ignore the alkaline-swilling troll.
>>
The only true logical way to control a mech
>>
>>16146785

> You want to somehow censor your own thoughts

If by censor you mean differentiate in a way that every single human who ever lived has probably done thousands of times in their life? Sure. It's not that you need to monitor your thinking super carefully or anything, just that you need to be minimally self-aware. The same self awareness every day people display when they distinguish between "I should get out bed/get coffee/keep working/whatever" and actually doing any of that.

> if it risks misunderstanding you

The entire point is that you train with the system, saving your neural patterns to the system so that it knows when you are simply thinking about something and when you actually want to do it. If there's risk there, it's so minimal as to be inconsequential and will only be an issue when a pilot gets really emotionally upset and no longer has complete control of themselves. For which you could check the statistics on crimes of passion to get a view of the risk I suppose.

> The TED talk you linked to makes my point.

I fail to see how, though you're welcome to clarify. As far as I can see though, the woman running it asks the guy to train the system to recognize his mental patterns going through a couple of tasks and within minutes he's capable of cogniatively complex tasks, with the computer able to identify what he wants. Which doesn't seem to support the idea it can't differentiate without several seconds of indecision without a clear answer to me.

> The first method to deal with this is to replace the pilot with an automated system.

That's certainly the more likely method yes.

> the second is to force the user to go through numerous checks and procedures.

Again, I'm failing to see why this would be necessary given that a computer can be trained to recognize the same mental patterns that we use to differentiate thought and action on a regular basis and execute based on that within fractions of a second; just like we do.
>>
An onahole. The Penis is a Mans True Heart.
>>
>>16146924
You really shouldn't bother. He doesn't care about truth. No matter what argument you come up with he'll find some nonsense to pretend like he's still right.
>>
File: latest[1].jpg (308 KB, 1280x720)
308 KB
308 KB JPG
>>16146927
I've actually thought about this before. Imagine if there was a mecha series about giant female mecha that were kinda sentient and the pubescent male cast had to insert their dick inside a hole in the cockpit to interface and control the mecha properly?
>>
>>16146964
EVA's cockpits are essentially wombs, but I like your idea better.
>hotblooded young man pilots a robot through sheer power of the D
>>
>>16146964

What happens if the pilot is a "one pump chump" and comes early? Does his unit just shut down? Or if the guy can't get it up? No start? Will it be like American porn, where all the guys are given viagra in order to ensure they get and stay hard for a given length of time and then afterwards there's a load of guys back at base nursing stiffies because the deployment didn't last as long as planned? Would the mechs be sentient and have preferences on length, width, curve or ball size? Would the males have to "court" the mechs by maintaining them, bringing them out on exciting flights for dates or upgrading their bits or doing paint jobs? What if a female wanted to pilot? would there be a possibility of a "trap" or "trap" preferring mech? So many questions.

>>16146985

I didn't see it, but it reminds me of Daimadler. The sexuality of the plot kind of reminds me of Gonna Be the Twintails too, even if the ideas have nothing in common.
>>
>>16146989
>exuality of the plot kind of reminds me of Gonna Be the Twintails too
Honestly, the basic premise of kinda-sentient giant female mecha piloted by a guy's dick would be a perfect fit for both an 80's hyper-sexual OVA, or a 00's/10's hyper-fanservice TV series. Get Fukuda to direct it (and get someone to make sure he doesn't blow the entire budget on stupid shit) and you'd have the masterpiece of ridiculousness of the decade.
>>
>>16140874
Fun fact: In VO, the remote control works by transplanting the consciousness into a live robot. When the robot gets destroyed, the body left behind becomes a vegetable. This was a plot point between the first and second games when a fuckload of kids were secretly conscripted to fight in the first game's events, only for them not to be good enough at videogames to make it through.
>>
File: 1478897962573.jpg (137 KB, 1280x720)
137 KB
137 KB JPG
>>16146924
Your point is that "technology will eventually solve everything... in the long term", doing so by "training" for an INFINITE number of pattern. Believing human are capable of what we call "inhuman focus" and will never have stray thoughts that ordinarily are just stopped by the time it take to do (or a interface that recognize the error and override your action).

My point, to reformulate it, is that until we reach this level of technology we will have a long period where Brain-Machine interface are good enough to "read pattern" from our mind, but far from being intelligent and reliable enough to make it a workable (or primary) way of piloting a mecha.
You are not going to train the mechs for 1000 methods of grabbing items of different mass, structures and consistency or to do every single things you'll ever ask a machine to do.
The goal of an interface is to allow you to do thing more easily, this is defeated if you have to micromanage and focus to prevent any misunderstanding.

By the time the technology get reliable enough, we will have reached the ability to make the robot sentient or wired our brain to have a direct feedback, point at which talking of "cockpit & interface" become pointless.

>the TED talk
Even assuming the technology improve, the entire video show just how much preparation you need to do anything and that the helmet will not recognize it instantaneously as if the "software was you". Every example of action are precise action that cannot be confused with each other.

So I will agree that it might find some very failsafe use in the short therm, like "select the target that you have detected (by reading both my mind and confirming through my eyes that this is the right/only target". But anything that have consequences would need to be locked behind the mechanical par of "trigger discipline".

This would include a mech basic movement, you can't live in constant fear that he cause a disaster by acting, or failing to act.
>>
>>16147049

> Your point is that "technology will eventually solve everything... in the long term", doing so by "training" for an INFINITE number of pattern. Believing human are capable of what we call "inhuman focus" and will never have stray thoughts that ordinarily are just stopped by the time it take to do (or a interface that recognize the error and override your action).

No, it isn't. My point is that technology is already being developed that can do that, even if it's in a nascent stage and that you don't need focus of any kind to control those thoughts; just regular human self awareness. You also don't need to train an infinite number of patterns, no more than a baby needs to; just train some basic ones and the computer, like your body is able to interpret based on those for any eventualities that come up.

> You are not going to train the mechs

You don't have to; that's literally the point. You have to train your brain to work the mech, not train the mechs computer through whatever number of variations on "pick up X" you want to name.

> if you have to micromanage and focus

Good thing that isn't what I'm suggesting despite your repeated assertions otherwise. Or do you have to micromanage and focus to control every movement of your body every second of every day?

> how much preparation

You mean the couple of minutes it took to put a helmet on and train two basic commands? Yea, however will people manage.

> the helmet will not recognize it instantaneously as if the "software was you".

Show me a baby who can walk out of the womb and tell the doctor there's no need for the customary slap and you might have a point. The brain needs training for everything, but familiarization literally enables it to become second nature and something you would actually need to concentrate just to fail at.

> you can't live in constant fear that he cause a disaster by acting, or failing to act

Do you constantly fear your own body will rebel on you or something?
>>
>>16146989
the mecha shocks/vibrates his dick to keep it hard, or maybe the mecha shoves something up the pilots ass to keep it hard.

Otherwise the hole gently sucks the pilot's flaccid dick so he's still in control, even slightly. Females can't be pilots. Fuck em.

Another idea would be for the pilot to stick his dick in the hole and then a mechanical tentacle goes up his urethra to slowly feed off his semen as a power source for the mecha.
>>
File: supreme commander.webm (2.96 MB, 1366x808)
2.96 MB
2.96 MB WEBM
a mouse and keyboard is all you need
>>
>>16147107

> The mech sucks the pilot off to keep them going

You had me.

> Or maybe it puts a tentacle down the dick to feed on semen.

And then you lost me. That was a good six seconds I suppose anon. Was it as fun for you as it was for me?
>>
>>16147129
I just want a kinky intimate way to control a mech
>>
>>16147107
>>16147129
>the villain mech of the setting is a super hardcore sadist dominatrix who's always finding new ways to exploit the power of the dick
We need to keep going with this
>>
>>16147151
>not another male
>both males fight each other in their robots, trying to make the other one cum first from all the impact and vibrations
>the one who comes first loses power
>>
>>16147151

There should be at least one lesbian villain give the whole "power of the D" thing. Especially if she's a love interest for someone on the protagonist side.
>>
>>16147160

> The main villain throughout the show isn't a sentient mech who thinks herself too good for any pilot and who becomes fascinated with the main character because she can't make him cum quickly like every one else, falling in love with him over time but seeking to move beyond humanity by "defeating" him sexually and proving herself superior
> Show doesn't end with the hero forcing himself in to the main villain's cockpit and fighting with her against some new threat that she served all along, finally forcing her to acknowledge him, both confessing their feelings while simultaneously defeating the enemy in a Love Love Tenkyoken moment
> Only instead of the Burger King, it's a big load of semen flying at the yonic villain mech as they both climax
>>
>>16140515
There is an old master grade guide that implies that, and it sort of explains half the reason why the learning computer was so important was because IT was the first one to well, do this, and it got a lot of high performance manuevers that were passed down.
>>
File: tfw you fart in a bus.jpg (31 KB, 147x142)
31 KB
31 KB JPG
>>16147167
>final boss is some giant devil gundam shit that defeats robots and their pilots and captures them to feed off their semen via tentacles
>>
>>16146964
Darling in the FranXX comes out later this year and is basically this, except a human girl is the cocksleev-I mean cockpit.
>>
File: ff13-shiva-bike[1].jpg (180 KB, 980x1050)
180 KB
180 KB JPG
>>16147191
imagine if your robot also had a waifu counterpart in the cockpit like this and you insert your dick and converse while in battle
>>
>>16147196

This worked better in Cho Aniki.
>>
>>16147216
well you would mount yourself on her in the cockpit in the missionary position. Like how Kallen does.
>>
File: ep1 fight suit.gif (1.17 MB, 320x240)
1.17 MB
1.17 MB GIF
>>16140535

I see your Rain and I raise you Domon
>>
File: 006.jpg (89 KB, 750x1098)
89 KB
89 KB JPG
Best mecha cockpit
>>
File: 1290032585827.jpg (137 KB, 1280x720)
137 KB
137 KB JPG
>>16147086
To be honest, I think you only have a layman's understanding of technology and didn't realize how much work and setup happen behind the curtain back to give you the illusion of magic.

A mind reading helmet do not make whatever is connected to it part of you own body, a mecha wouldn't know magically know how to move its limbs and where to put his foot just because you thought "go this way".
So your smartass:
>do you have to micromanage and focus to control every movement of your body every second of every day?
Miss the point that you do not control your body by having an algorithm look at crazy electric signals and be programmed to only do a set of action if it recognize a fuzzy pattern that can't be confused with another.
It took a nervous system with direct feedback and millions of years of evolutions for us to walk and YES you do micromanage your movements unconsciously. Your brain is just amazing and don't bring most of the work to your consciousness, that's why you have involuntary reflex.

When a monkey controlled a robotic arm it was with a hardwired link, by doing so it was an extremely adaptive brain that took the stress of learning to what signal do the arms do what, not a computer chips "guessing" what those signal means.
Current exoskeleton read the nerves for the same effect.

The TED talk isn't about showing their failures so they didn't ask anything tricky like "focus on not doing this" or "focus on the possibility that you might do this but don't".
Take this as a proof the TED talk is really nothing new to me.
https://www.wired.com/2016/11/used-mind-fly-plane-around-seattle/

>You have to train your brain to work the mech
btw, didn't I say interface are supposed to make it easier to use, not harder?

Again: brain-reading is fundamentally flawed because it is a read-only interface working from exponentially fuzzy data on a ever-changing brain.
The more actions you train to do, the less a computer will be able to distinguish them.
>>
>>16147151
>>16147160
>>16147163
>>16147167
>>16147179
https://www.youtube.com/watch?v=drvAZ7e_fAw
I think I got your climax
>>
>>16147126
Well, and AI for your troops.
>>
Even if it is possible to do a neural interface that works, what advantage does it give you over a person just using an ordinary controller and a bit of "fly-by-wire"? In ranged combat it's certainly a huge waste of training. For melee combat I suppose it might give you a bit of an edge, but you would have to train bloody hard to take advantage of it. Perhaps it could be something limited to the most elite troops, chosen at an early age of course for plasticity.
>>
>>16148870
Assuming the machine is fast enough to take advantage of it, it means faster time to shoot and the ability to have more fluid/human movement without needing to simplify the controls which can result in inaccuracy or clunky movement. It's not that controlling a gun with a button or joystick doesn't work, but there are sacrifices to be made in terms of control. Do you want speed or accuracy? Finely grained control or fast traverse?

But depending on the scale of the machine, it's role, and how it's meant to be armed, there's probably only a small niche where neural control makes more than a small difference.

The closest real life analogue I can think of for neural control would be the helmet mounted gunners' sight for attack helicopters. Instead of needing to aim using a joystick and looking through a TV display hooked up to a camera mounted on the gun itself, the gunner's helmet has a tracking system where the gun moves according to the gunner's head motions. Of course, it's not actual neural control, but it offers similar benefits and frees up the operator's hands for other things.
>>
File: 3Zenobia.gif (21 KB, 390x415)
21 KB
21 KB GIF
>>16143139

This works.

>>16146407

Ah, just pointed out the most logical pilot scheme

>You won't
>Because it's autonomous

>>16146785

Wait, is one of you implying that it would be easier to control a mech using physical controls than your mind? Because that is entirely not true.

Pic related, it has a crew of two, uses a physical interface and still takes two fucking years of training to use.

>>16148870

If you have an ordinary controller with fly-by-wire you shouldn't be using a mech to begin with.

>>>/k/
>>
>>16148956
It took a lot longer than two years for you to learn how to control your body effectively. More importantly that's actually a body you evolved with- the adaptability of natural brains to new toys is impressive but it has limits.
>>
File: 615.png (13 KB, 223x200)
13 KB
13 KB PNG
>>16143594
WHAT IS THIS FROM I NEED TO KNOW
>>
>>16140535
>>16147298
Why does this turn me on?
>>
>>16141344
I second this
>>
>>16143594
>>16149288
Gun Sword
The girl in qurestion was a former prostitute
>>
>>16149362
thank you good sir
>>
>>16149327
Don't panic, this is normal.
>>
>>16147298
I wish I had a ass like that in high school
>>
File: 1469252233224.jpg (165 KB, 1200x1200)
165 KB
165 KB JPG
>>16146477
>Your own body is basically just a suit to protect the brain that it learns to adopt to over the years as is.
That got a little meta for me desu
>>
A bodysuit that controls a mecha through muscle movement and a HUD display that responds to eye movement and voice commands.
>>
>>16141344
Semi-seconding this, but a step further. I really like how Dai-Guard functions. One main pilot in control of basic movements and primary functions, a second pilot in control of power adjustment and the finer movements that supplement the bigger basic movements, and a third pilot in charge of radar and other non-movement based systems.
>>
>>16140331
And if you are?
>>
File: p11.jpg (221 KB, 636x375)
221 KB
221 KB JPG
Sleek, simplistic effective design. Able to with stand direct blows for short periods, effectively draining is power cell over time. Stable canopy glass. Easy control set up
>>
File: giphy.gif (151 KB, 640x400)
151 KB
151 KB GIF
>>16149807
>>
>>16148956
>Wait, is one of you implying that it would be easier to control a mech using physical controls than your mind?

Because it is. Mindreading can't be used until we reach a technological level that make the mech autonomous and sentient enough to guess what you meant/need and discard bad data.

Your picture is fiction, see >>16148531 and https://www.wired.com/2016/11/used-mind-fly-plane-around-seattle/ for a real attempt.
The technology won't improve for fundamental reasons explained before. The more numerous the orders are and the more precise they are, the harder it become to separate them from each others.

This is the equivalent of trying to pilot by handing notes of badly written and progressively harder enigmas to someone else who then try to guess what they mean and then pilot by himself.
If you want real control you need a direct brain-machine interface with feedback so the response time is at its shortest and order are not guessed but as clear as 1 and 0.
You'll still need an onboard computer but it won't need to be sentient.

>>16148936
Engineer discovered designing the F35 helmet that it is easier to read the position of the eyes and guessing your intent from muscle movement for tracking than reading the brain.
That's why we need a direct neural, the feedback would change everything. Reading the whole brain is only useful to track the overall burden of the pilot and improve other interface.
>>
>>16149677
The world would be a better place if everyone had great ass.
>>
File: 1259567207057.gif (13 KB, 150x106)
13 KB
13 KB GIF
onboard AI is needed
>>
>>16152481
That probably goes without saying. Even today's jet fighters have computer-assisted controls.
>>
>>16139551
No way to controll a robbot inside a cockpit is realistic when it is human shaped.
>>
File: 1328890606523.jpg (80 KB, 600x744)
80 KB
80 KB JPG
>>16153549
Why not? We all agree that onboard computer will be necessary anyway as you can't trust a pilot with the balancing or the math behind everything.
So it's only a question of what sort of interface inside the cockpit is more efficient.
>>
>>16150685

> That's why we need a direct neural, the feedback would change everything.

This one line seems to imply the entire argument was kind of pointless, because I was trying to discuss direct neural interfaces all along. You appear to have thought I meant the computer just reading the thoughts of the pilot as they sit in a regular cockpit with no direct feedback of any kind or something and the pilot only experiencing the fight through monitors with a computer acting as a liaison between the pilot and the mech, relaying orders the pilot gives to the unit.
>>
>>16156216
Yeah, the only problem with that is it involves some rather invasive surgery. If you want to read brain signals wireless might just cut it, but putting signals back in wirelessly seems very unlikely.
>>
>>16156424

It doesn't need to feed directly from the machine in to the brain and there's a good possibility that sensation could be passed through the skin in to the nervous system and up to the brain. The cockpit could be use something akin to the LCL in Eva or simply be a chair with sleeves to insert limbs in to with conductive material lining the insides and able to pass information from the machine to the nerves through skin. There's no reason that information couldn't permeate the skin that I know of.
>>
>>16140505
ing zero's cockpit is pretty much this, with Zero A.I taking care of the rest
>>
>>16149807
>>16149808
>Sheesh roger, you will going to learn that you're not a Skywalker?
>Order a guide
>Start Again

fuck, we need a Space quest remaster ASAP
>>
i think MWO Pretty much nailed the concept
>>
>>16156424
>invasive surgery
You don't say...

https://www.youtube.com/watch?v=zFIHYJAp2rc
>>
>>16139551
Well, there's mental controls being developed for cars right now for braking/accelerating. Would be easy to apply that for mech acceleration control while joysticks are used for arm controls.
>>
File: Sexy Bionic sort of.jpg (396 KB, 2755x2292)
396 KB
396 KB JPG
>>16156216
...I was ____________________________ this close of writing "unless it's a misunderstanding"... in one of my answer.

However taking the TED talk and reading the brain from afar as example is still objectionable. To reformulate again the problem is to remove the noise and make sure the brain/machine interface can recognize an error and recognize a context if the order stretch too far to be considered a direct-feedback.

>>16156529
The problem with reading the skin is that you are reading nerves, it work fine for single signal answer but if we are talking of order like "go forward" it would not exchange enough information to, say, avoid something on the way that the onboard computer cannot understand.
Also I INSIST you DO face the problem of involuntary movement (unless the computer can read enough context and ignore the order)

Direct-feedback only make the computer much much simpler because you can reduce exponentially the exchange to binary output. It also make it easier to discern involuntary movement as noise.
Defining context where the computer have to "obey order" or "don't obey" will still be the most difficult part.

You don't want the computer accomplishing a move and jumping when you where only mentally preparing yourself to do the full sequence and feel your body react as if.
The direct-feedback only facilitate to set up a "nogo" condition within the decision loop.
>>
>>16157562

> the problem is to remove the noise and make sure the brain/machine interface can recognize an error

I've no idea why you think this is such a big problem, because even after 2 minutes on stage using relatively cheap first generation technology there wasn't a major issue with errors. That isn't to say there would never be problems using the system, only that the kind of problems you are worried about can obviously be addressed through time with the system.

> The problem with reading the skin is that you are reading nerves

You could be, but I wasn't suggesting you were; just that something like the wireless headset be used for reading the pilot's brainwaves and something else be used for transferring feedback to the pilot. The input and output do not have to be the same after all.

> I INSIST you DO face the problem of involuntary movement

Again, I've no idea why you think this is some major issue that may not be overcome. Your brain doesn't feed your body false information because your brain trains your body over infancy to recognize the type of commands it issues. You do not have to train it to recognize the difference in every single variation between "thinking about X" and "want to do X right now". "Thinking about" thoughts have a particular footprint to them, regardless of what you're thinking about and the same with "do" thoughts, so you train enough that the system becomes familiar with the difference and able to differentiate them in all variants going forward. At that point, involuntary movements will be a really, really minor issue at best.

> Defining context where the computer have to "obey order" or "don't obey" will still be the most difficult part.

Hence: training.

Also: yes, Aimee Mullins is very sexy. There's a reason she's a model despite being a double amputee. She also has something like a dozen different "legs" of various kinds, including ones that make her taller as well as running legs, modeling only legs etc.
>>
>>16160281
>I've no idea why you think this is such a big problem
I would say that's because I couldn't get you to get my point. So I'll try another way.

First for the sake of clarification my first post >>16146407 was taking the extreme case of a mech(plane) solely controlled by a brain with the intent of reducing the reaction time. That and keeping control is mutually defeating. So it was as much about mind Reading than Brain Interfacing. pict was from Macross Plus.

About reading nerves, the point is that a nerve is infinitely easier to read without errors, they are the only part who don't receive easily confusing data. Reading the brain is like reading a constantly evolving mass of billions of cells that are not guarantee to give a reading clean enough for useful purpose (and that's even if it didn't changed during your entire life).

>You do not have to train it to recognize the difference in every single variation between "thinking about X" and "want to do X right now".
You do, because you are reading the brain, not the nerve who don't react to the "thinking about X"
Especially when "doing X successfully" especially require to ignore false negative/positive for a while (the problem of G-force and joystick).

>"Thinking about" thoughts have a particular footprint to them
They don't at the level you believe they do, we are not robot or clones.
We need to implant electrode to get to guess 70% of the time what word someone thought about. Words being the minimum level you need to get right at 99%.
The TED talk did not have the machine recognize the brain cell for the concept of "cube disappearing". They told the machine to make a cube disappear if it noticed any pattern that looped and fit with what they barely recognize as satisfaction statistically speaking.

Reading the overall brain will be useful,
But the reading method is fundamentally limited and counter-productive for a reliable "main" control system.
>>
>>16160636

> I would say that's because I couldn't get you to get my point. So I'll try another way.

I'm pretty sure I did and just don't agree with you on it being a problem. The issue isn't that you're failing to identify a problem, it's that your identifying it and I'm going "but that's obviously not an issue".

> You do, because you are reading the brain, not the nerve who don't react to the "thinking about X"

This is demonstrably enough, because in a linked demonstration using current technology someone's patterns were measured using one thought and a few seconds later he was capable or ordering a different thought using the same registered patterns because those patterns are similar enough to have more than one application.

> They don't at the level you believe they do, we are not robot or clones.

Why would we need to be to have thoughts of a specific kind have similar, detectable elements?

> We need to implant electrode to get to guess 70% of the time what word someone thought about. Words being the minimum level you need to get right at 99%.

Words are a lot more specific than thoughts, which is probably why people often struggle to put their thoughts and feelings in to words.
>>
> The TED talk did not have the machine recognise the brain cell for the concept of "cube disappearing". They told the machine to make a cube disappear if it noticed any pattern that looped and fit with what they barely recognise as satisfaction statistically speaking.

The computer was what made the animation, but the animation is based upon his input and that input is (a) coming from brainwave detection and (b) based upon detecting the user thinking about what he wants the cube to do. Even if that isn't reading specifically the thought "pull/hide the cube", but instead upon a pattern of some kind (which isn't indicated in the talk) then that's input is obviously compatible with control of external objects regardless so I still fail to see the problem.

> But the reading method is fundamentally limited and counter-productive for a reliable "main" control system.

You have yet to indicate why it would have any kind of limit, only that you think it exists and direct control of a machine is not a good idea because it (apparently) exists.
>>
File: 1263373396122.jpg (62 KB, 720x480)
62 KB
62 KB JPG
(1/2)
>>16162348
>You have yet to indicate why it would have any kind of limit, [...] is not a good idea because it (apparently) exists.
Clarification again, I explain how that "limit" is more an exponential cost increase (per unique detectable pattern) that technology can't match.
I gave you a link related, we have prosthetic, and I explained why it won't "get better as technology evolve"
https://www.wired.com/2016/11/used-mind-fly-plane-around-seattle/
So all I have left is boasting that unlike you I understood the TED talk, they didn't hide the limitation but I'm afraid you seem to have replaced it by an idealized vision.

>This is demonstrably enough, because in a linked demonstration
To me, all of the demonstration linked so far only showed computer incapable of reading more than one pattern and only after being told what to associate this pattern with.
Up to now it took direct link & electrode to start reading (instead of guessing) what you are actually thinking.
Hence why I put aside neural-interface which unlike the other will get EASIER the more data you try to read, thank's plasticity.

>(b) based upon detecting the user thinking about what he wants the cube to do
This do not match what I watched, read and what I know the tech to actually do.
The entire point of the calibration, "training" as you say, is to have the scanner detect a new pattern (after telling the user what to think about) to set create a "word"+ a range from 1 to 10.
At no point do the scanner understand what the pattern is or if it is similar to another, like "have the cub match the background color" or "cease to exist".
Unless we can predict how the mind work with an amazing computer, you won't get at that level of detail and you need to.

For clarification, while it merge in part, there was a different logic behind me saying a brain-interface need to read context to avoid involuntary movement.
>>
File: 1368595717935.jpg (127 KB, 720x540)
127 KB
127 KB JPG
(2/2)
>Why would we need to be to have thoughts of a specific kind have similar, detectable elements?
We don't need it, it how it is. You don't have neurons dedicated to activating afterburner, only instinct are kind of like that.
The plasticity of the brain make people very different from each others as individual and culture: e.g: a culture with words for 20 shades of white.

>Words are a lot more specific than thoughts, which is probably why people often struggle to put their thoughts and feelings in to words.
And this is precisely what you need to exchange good information. A concept you've mutually agreed to name and restrict the meaning to prevent any misunderstanding. (we are not newtype)
http://www.dailymail.co.uk/sciencetech/article-3873980/Forget-typing-Computers-read-MIND-convert-thoughts-text-way.html

For the sake of demonstration, let's assume thought converted into order to give:
<forward, 30° left, full speed>
But you cannot guarantee a full word can be caught within the margin of error and have an unique pattern (even with training), or can be recognized in time before the next order come.
You don't want your mind-reader to accidentally catch <forward, 30° left, null speed> or <.....left, full speed>
A solution is to repeat the order and samples, but in our case this would liken to concentration, time and block opposing orders during the sampling.
Another solution is to accept more "thoughts" at the cost of precision <forward/front/onward/toward, .....etc>, in our case that's akin to reading more of the brain, widening the "pattern".
Lastly you ask more of your attention for an action when the goal of a interface is to do the opposite.

Physical control address all this by trying to make physically impossible to be wrong, switches and inertia to keep them valid. (prosthetic) equivalent of nerve/neural-link get closer to this easier because the feed-back teach the brain to only replicate the signal associated and nerves have cleaner signals.
>>
File: 082[1].gif (1.46 MB, 398x300)
1.46 MB
1.46 MB GIF
like this
>>
>>16139551
ps3 controller
works good enough in call of duty
>>
>>16143477
Build fighters works like a videogame (Gundam Breaker to be more accurate)
You build a gunpla and map individual movements(attack animations) to it.
Would explain all the reused footage in a really meta way
>>
File: TOY-GDM-1418_02.jpg (73 KB, 550x800)
73 KB
73 KB JPG
What if you could control a robot like a doll?

like have it mounted in the cockpit and and move it's legs and arms with your fingers like playing with a toy. To fly you press pedals and buttons.
>>
The best way to control a mecha is to have it made as simple as possible when it's in action, but give the pilot more access when maintenance is needed.

1.Have a pilot and have an AI onboard

2.The pilot is the human element, and controls the mecha through preprogrammed actions (walking,dashing,shooting etc.)
But there could also be a psycommu type of thing that reads their mind to refine these actions when needed.

3.The AI is there to work with the pilot and also make adjustments and regulate other systems as needed, but the pilot can handle this stuff if they choose to

>>16146819
It looked so fun, I can't wait for their return in S3
>>
File: latest[2].jpg (30 KB, 245x400)
30 KB
30 KB JPG
>>16164268
Controlling should be like having a haro on board. You converse and tell it to do certain things.
>>
>>16164263
That sounds really interesting, but unless the cockpit had a really good G-dampening system it would be extremely difficult to do.
Imagine trying to hold on to some action figure while being slammed around by high G-forces, you might break the doll controller that way.

But this gives me an idea, maybe you could split the doll in half vertically so that instead of joysticks you have 2 halves of a doll to manipulate.
This would put the pilot in a less awkward orientation in the cockpit and be less risky.
>>
>>16164269
yeah that could be perfect.
Each pilot would have a personal unit so that overtime it would learn their combat habits making the process more efficient
>>
>>16164271
The figure would be mounted in front of your chest like a steering wheel. And the doll would be made of steel and stuff to not break.
>>
Fortified suits do have the right about putting the camera feeds on the equivalent of a helmet rather than screens. In terms of cost effectiveness the only reason why you'd use video screens in a cockpit is so that people not wearing anything special can still use it. The price of covering every square inch of a cockpit with a screen for a panoramic view doesn't make sense compared to just putting it all in a pilot's helmet. If that thing gets broken then the pilot is likely to not be in any shape to continue piloting anyway.
>>
>>16139551
Macros. Lots and lots of Macross.

We already have bipedal robots with mobility comparable to humans. Just scale that up and make it pilotable via, hell, console gamepad. For direct control some sort of waldo http://muppet.wikia.com/wiki/Waldo_(remote_manipulator) will work just fine.
So, valkire control scheme, really.
>>
>>16139559
So Battletech then.
>>
File: Cockpit.jpg (1.06 MB, 2986x3140)
1.06 MB
1.06 MB JPG
>>16164344

Here is art from an /m/ related project I am working on. Would this sate your autism?
>>
>>16164314
Huge panoramic screens like in gundam might be excessive, but a collection of normal sized screens like in steel battalion might be appropriate. That way you make it easier to separate sensor feeds from guided weapons, targeting pods, GPS maps, instruments etc. Head screens are something which we've only just started being able to do in real life so it's hard to say what the advantages and disadvantages are, it might be that screens are more comfortable to use over extended periods.
>>
>>16164362
Seems reasonable enough, deliciously retro as well.
which project?
>>
File: 1494579717670.png (233 KB, 524x541)
233 KB
233 KB PNG
>>16139551
Darling in the Franxx just made the best cockpit because you get to see a cute anime girl's butt while you pilot
>>
>>16139559
Pure cringe
>>
>>16143139
Command was a nice game for the psp. The journey from Eath to the Dimensions and finally to the home system of the Bydo were well worth it, even though the interactions between commander and pilots are basically nil.
>>
the same way i play my mmos
>>
>>16139551
remotely.
>>
>>16139987
AC units break the sound barrier in all directions in combat, not sure how exactly a human can survive that shit.
>>
>>16167930
with bad posture and drugs?
>>
>>16168159
If you lock human body down, it can withstand high G forces just fine, althou here we encounter interesting problem of users brain slushing around skull. A GitS style cyberbrain might be in order.
>>16168172
And macros
>>
>>16168774
It's not just the brain, eyeballs can also be a problem as with blackout and redout. No amount of bondage is going to help you there.
>>
>>16169350
You can bypass this problem by using direct brain interface, as they do in AC verse anyway. What's hard limit for humans anyway, 60G?
>>
>>16169486
trained fighter pilots wearing pressure suits to counter g forces will pass out within seconds when performing maneuvers that result in up to 10 g of sustained continuous force like staying in a really hard turn or circle

in various circumstances, untrained human bodies have tolerated up to like 100 g but only for fractions of a second, such as in disasters and crashes
>>
>>16166627
What would you suggest?
>>
>>16169664
So, theoretically, if we eliminate squishy human body, reinforce brain by putting it into cybershell, Major style, we can have 'pilot' stay functional at speed AC4 frames are capable of achieving. This is of course, stupid solution, as here AI or AI reinforced telepresence systems make more sense but hey.
>>
File: 1280923253167.jpg (56 KB, 640x480)
56 KB
56 KB JPG
>>
>>16147148

Does Darling in the FrankXXX's control method of a male "riding" a female do anything for you out of interest?
>>
File: 1343946357991.jpg (151 KB, 952x716)
151 KB
151 KB JPG
>>
>>16146964
>literal cockpits
FUND IT
>>
File: Frame Jehuty.jpg (54 KB, 576x809)
54 KB
54 KB JPG
>>16174876
I heard you wanted cockpit
>>
File: 1515916422701.png (814 KB, 739x1080)
814 KB
814 KB PNG
>>16147148
>>
>>16177098

Is the monitor powered by bio-electricity from an anal plug or something to say there's something running from the helmet down the spine to her asscrack? Does the pilot suit just have handles hanging off the hips all the time?
>>
>>16179393
All we know is that you gotta mount a bitch to go full humanoid.
>>
>>16139551
Well, that would depend entirely upon whether the mech was designed to fight in a zero gravity environment or a planetary surface battle zone.....These are mutually exclusive.....one will not work in the other.....
>>
>>16179393
You're thinking too hard about it
>>
>>16180695

But that's why it's fun. Same as wondering if the monitor on the back of the helmet means if the guy could watch porn while he fucks her. Or if he could live stream her face on to the back of the helmet and watch her face as he does her doggy. Or if the suit getting knocked around by an enemy would cause him to accidentally insert.
>>
>>16180670
i must have a weird penis , cause i barely get any pleasure from doggystyle and nips are even designing cockpits around that posture!?
>>
>>16181338

Probably because it's one of the least awkward sexual positions to design a cockpit around. Missionary for instance would be harder, because the pilot's viewlines would be blocked by his partner and he'd have to constantly crane his neck to see around them.
>>
>>16181347
you sound right, leave it to the japs to take away the fun even in sex...
i think you could do something better with some missionary variation and still have room for stupid control handles integrated in the female pilot suit, without the need for a neck-monitor...anyway hows that rule 34 coming along?
>>
>>16181358

I'm not really seeing how you'd make a missionary position cockpit without changing the position itself to the point it's no longer really a missionary position. That said, I do expect that at some point near the finale the main unit in Darling will evolve or change in some way and get a new, more romantic cockpit block. It might not be missionary, but it'll have the two pilots in a more similar position probably.
>>
>>16180690

Why would they be mutually exclusive? Mutually specialised I can certainly see, but exclusive seems a bit much.
>>
>>16156646
Pretty much this
>>
>>16183269
AAAAAAAANd is still great because its pretty much the idea of someone who has said mech for a looong time
>>
File: vzrtk6k7ez5mfqxhkxur.jpg (70 KB, 800x566)
70 KB
70 KB JPG
With some cardboard and string, obviously.
>>
File: NEET life.webm (1.45 MB, 512x384)
1.45 MB
1.45 MB WEBM
You all aren't thinking luxuriously enough. Haven't you ever wanted to pilot a giant robot from a comfy velvet divination room?
>>
This show really does have the best cockpits
>>
>>16185558
>>16185564

Source? Seems like a fun show.
>>
>>16186771
Both are from Ryu Knight
>>
File: 1336826900517.jpg (88 KB, 1366x768)
88 KB
88 KB JPG
I'm not sure why the Entry Plug of Evangelion have such a great appeal.

We've seen long cockpit before, sleek one, stylish one.
If someone want to try to answer that, I'll actually read it.
>>
File: 1445099111255.jpg (783 KB, 1400x1334)
783 KB
783 KB JPG
as payment for this >>16187885 answer, have this.
>>
File: cowboybebopinhismech.jpg (278 KB, 839x472)
278 KB
278 KB JPG
super advanced and spacious




Delete Post: [File Only] Style:
[Disable Mobile View / Use Desktop Site]

[Enable Mobile View / Use Mobile Site]

All trademarks and copyrights on this page are owned by their respective parties. Images uploaded are the responsibility of the Poster. Comments are owned by the Poster.