Humanoid robotics are pretty much the only thing I get excited for but it's also seems to be the area of robotics with the least funding and there are so many difficult challenges to overcome that I feel like nothing interesting but some service robots with stiff movement and chatbot tier AI will happen in my lifetime.
What do?
There is no practical application.
It's easier to just pave everything or use flat terrain most of the time. Or just make more rugged/inefficient wheels that can get past obstacles.
Legs and limbs took a very long time to evolve to the optimal state they are in now. It would be very hard to recreate them using only human brainpower.
Basically there is no real economical reason to develop them. It's an extremely complex problem and we have the great ability to alter environments as humans. So we can just alter environment to reduce obstacles and use wheels, which are simplistic.
Get a research grant for humanoid robotics research
Road paving is expensive you underestimate the load it would have on our infrastructure to have more wheeled vehicles around.
Also legs being evolved does not correlate to complexity. Computers we're originally a hired occupational role.
dont worry OP, self driving cars will pave the path for autonomous humanoid robots as well.
the advanced radar tech and will make it much more practical for legged robots
they have the walking down, its a simple pattern of movement.
in fact many people don't know this but we don't actually use our brains to walk, we use our spines, a circuit called a CPG (central penis generator).
the brain sends the primary signal, then the spinal cord takes over.
the issue with robots is sensory, they cant compute the information fast enough for a human-like reaction time that extends beyond stationary tipping (ie tripping, pushing, slipping, etc).
so with the radar tech used in self driving cars which can react inhumanly fast and employs advanced trajectory/prediction algorithms, our metal-legged friends will soon one day be walking among us, and not bumping into us, tripping, falling, or slipping.
:)
*central pattern generator
no homo
I don't see the value of creating bipedal or even Quadruped robots. There is an immense complexity to limb-based locomotion. Wheels are simple and fairly efficient.
Yes, it's expensive to pave everything, but even with limbed locomotion you'd probably want most things paved or very flat.
The complexity of legs involves a lot more than just the kinematics and balancing. It involves recreating something similar to muscles. Most current methods use servos, which in comparison to muscles are shit.
I don't see a point behind "researching bipedal" stuff without much better muscle-like technology being available.
Which is an incredibly complex task.
Basically wheels, rotors, and jets are fucking simple and easy compared to developing limbs.
I just see zero practical application besides creating neat tech demonstrations until we have much further advanced technology.
Militarily they are pointless as the cost is too high when every future battlefield will basically be full of autonomous drones blowing shit up that cost next to nothing.
>There is an immense complexity to limb-based locomotion. Wheels are simple and fairly efficient.
There really isn't. Don't just assume that because animals/humans do something, it automatically has to be super complex or advanced.
Okay, I can see cars fucking rolling around smoothly
Yet the best humanoid robots are jerky servo shits that a 5 year old could fuck up in a fight.
What's the point exactly behind creating million dollar jerky robots that suck ass versus a 2k wheeled robot?
Again, I'll just use the fucking example of no financially viable humanoid robots fucking EXISTING.
Humanoid robotics is stupid. We as humans need to get over our narcissistic obsession with making robots that look like us.
Robots should be designed in a way that best facilitates their function.
>Central penis generator.
Top kek.
better question :
how come japs robots are still the same utter shit despite them being developped for 20 years at the very least ?
>This
I agree with the narcissism thing, and similarly I hate the way that our concepts of alien life always seem to resemble humans in most cases.
However there is a lot of sense in designing robots like humans or animals (like Boston Dynamics' Big dog for example)
Evolution has taken millions of years to perfect the mechanisms of our bodies, so mimicking these proven systems gives us a head start.
There's also the social aspect of wanting robots to be accepted by humans as non-threatening and "just like us"
>There is no practical application.
The only practical application is the ability to use things that humans use and be able to move freely between those jobs/devices.
tl;dr sexbots are needed
>our concepts of alien life always seem to resemble humans in most cases.
That is specially for entertainment media so that humans can related to it. If you get too "alien" then the movie sucks for plebs and brainlets and doesn't make money.
Because the breakthrough that made American bots great was fairly recent and hasn't been copied to everything yet.
There was a Japanese entry to the DARPA grand challenge right?
I don't remember it doing very well.
>the breakthrough that made American bots great
What is it?
Americans themselves. Easy to manipulate and steal money from.
bump
Are you pretending to be retarded? You better not be.
The Animatrix has a great sequence showing how as the robots built robot descendants and improved their technology they made them resemble humans less and less, to the point that their bodies looked more like robot squid insects and their psychology was completely different.
Humanoid robots are pretty neat, but I don't understand their application. Humans are built to be pretty general purpose; really, the only thing we have going for us is our intelligence. Modeling a robot after our firm seems pointless.
I've been binge watching ghost in the shell and I will politely disagree with your first statement.
most robots will use wheels or treads because they're cheap, simple, and can support more weight with less energy
paving is cheap, even illegals in front of home depot are capable of doing it. Right now you can buy some quikrite and do it yourself. The main issue is ramps, but not really given that ADA compliant ramps were made for wheeled vehicles (mobility scooters) in mind
GiTS is an animated movie, not a manual
As technology progresses and we eventually create a humanoid robot with intellect close to or rivaling our own. There are 2 possible outcomes.
1) We see this new intellect as a foreigner. Something to be hated and feared because it is different than us. It's human nature to hate things that are different. Too different and alien and it's ok, but once you get an AI that mimics or rivals humans it creates a psychological uncanny valley effect. This path leads to the robot war.
2) This new robot is something that become a sorta role model. Something for humans to strive to become more like. We may discover ways to technologically augment ourselves to improve our minds and bodies. Eventually the line between artificial intelligence and human intelligence will become so blurred it'll become moot. This is the technological singularity.
This is basically the plot of the original Ghost in the Shell movie. While humanoid robots may not be useful or important now, it's important for the future of mankind.
In case it wasn't clear, humanoid robots are a step in blurring that line between man and machine.
the show isn't nearly as good as the movie
sexbots... porn has an infinite amount of money behind it... make a decent sexbot and you will have all money you ever desired.
Their was a major gait breakthrough in bipedal robots. Boston Dynamics' Petman and ATLAS both walk unstably, that is in each step they are not completely balanced and have to catch themselves before they fall on their face.
This leads to better variety of terrain they can handle and better speed.
>As technology progresses and we eventually create a humanoid robot with intellect close to or rivaling our own. There are 2 possible outcomes.
>1) We see this new intellect as a foreigner. Something to be hated and feared ... robot war.
>2) ... the line between artificial intelligence and human intelligence will become so blurred it'll become moot.
This is completely idiotic. You can't think outside of what you've been shown in fiction, and fiction is driven not by rational analysis but emotionally satisfying stories.
You've just listed two kinds of popular stories that benefit from the inclusion of fantasy robots: a war story, and a miraculous transformation story. Other good fits are a story of a lonely outsider seeking acceptance, adventuring in strange lands with exotic companions, and dealing with a monster.
What makes a good story has absolutely no bearing on how things work out in real life, aside from the cults that spring up.
Robots are devices, not life forms. The first thing you should think about is: what do its builders want it to do? The next thing is: how competent are they? After that: who else will have the same capabilities at the same time, and what will they build?
The one thing you should absolutely spend no time entertaining is the idea that robots will be basically people inside, especially that one which presents the appearance of having typical human emotions can be trusted to be moved or limited by them in the same way that a human would. The presentation of human emotions by a machine must always be assumed to be a (literally) calculated deception.