has robotics hit a ditch?

[Killerwolfbot]

In the 1970's (before I was born) There was a huge boost in the popularity of robotics. Today we are at the point were the technology easly avalible to anyone who wants to try to build there own little robot friend. So is it me or does anyone else feel like robotics should be more advance then we see it today? Theres no Irobots out washing my car or walking my dog and I for one feel like there should be. Not becasue I'd rather sit around and watch TV all day but becasue I love robots and I personally don't see much improvment over the years. I'm sure there have been plenty of advancements that I am just to young to recognize but the same basic robot remains. You buy a kit, whats it do? Dance, play soccer, "fight" if you can call it that. Even the newest of robot kits still walk sideways. Not to mention there all around 1 foot tall. It seems like bigger and better robots could be out on the market right now. So why arn't we seeing them? Or am I simply expecting to much?

Also has anyone ever tried to build there own humanoid robot from scratch? How did it go?

[Gertlex]

The answer that comes to mind for me is the cost of big stuff.  The Darwin-OP is a $12000 bot (biped) and somewhere under 2ft tall.  To make a household bot that is worth its price tag of tens of thousands of dollars, it needs to be able to do a lot of things well, and that's perhaps the real, huge challenge.

There's lots of people working on complex stuff (e.g. the stuff done in research labs/universities), but it seems a lot of stuff is just being perpetually redone without progress made...

That's what I got from 5 mins of thought on your questions

(Also my gf would advocate that if you/family don't want to walk your dog, you shouldn't have one, but that's just nitpicking! )

[Soeren]

Hi,

A biped is not very efficient, as lots of power will be consumed to just keeping it upright.
That's why most mammals use 4 legs for motion/balance and if more people get over the Creation aspect and start thinking in terms of problem solving (i.e. how to best perform a certain task), I'm sure you'll see a faster advance.

That said, i think you overlook the fact, that automation is all around us in specialized applications; like in clothes washers, µwave ovens, sewing machines and such boring home equipment.

It will be a long time before you see androids walking freely (or at all) - I don't think it will ever happens, mainly due to the power inefficiency - don't consider movies as an authority . it may be easily done in CGI, where you can neglect the physical properties of the real world, but in the real world, robots will stay "inhumane" 

Ask yourself... Why do I want human-like robots.
If it's something about social interfacing, perhaps you should go out and meet some flesh&blood people, as they're already capable of the interfacing part.
If you just want someone you can control, the dog you mentioned should do.

Why would a car washing robot need two legs?
And why would a dog walking 'bot ?

Exchange the dog for a robot dog and you won't even have to walk it 

[corrado33]

I think one of the biggest problems with creating the robots we want is that we don't have a good means to power them.  Batteries?  Too heavy and not powerful enough (without spending a fortune).  Wires are just annoying all around (especially for a mobile robot).  And if you're going to have wires connected to it why even make a robot, why not just connect everything to a computer and control it that way?  (Yes that's still technically a robot)

And we don't have mini mobile nuclear reactors yet... so....

I think everyone that starts with robotics wants to build a cool mobile robot that does something.  No one wants to build 20 different robot biped legs to see which one is the best design.

Personally, I'd love to build a robot foot.  I personally think that we could build better bipeds and quadrupeds if we had a controllable foot.  I mean think about it, how much do you use your foot for stability/traction/sensory information?  Even some of the most (ok I haven't looked in a while) bipeds still have flat or curved feet.  I mean, it could even have implications for prosthetics.  (It'd have to be really good/solid though.) The question though, is why?  Like Soeren said, why build a biped if something else is much better suited for the bot? 

One reason I CAN think of for building legged bots is traversing rough terrain.  Tracks and wheels don't do well over large rocks unless they have awesome suspension systems (think mars rover). 

With that said, I'm pretty much making all of this stuff up, so I'm probably wrong, but that's how I see it.   

[Killerwolfbot]

3rd time tryin to reply XD         i see the human body as a multitool. I(well probably not me) can program a humanoid robot to do a verity of tasks without having to replace major componets. also like you said im new to robotics and want to build something cool.

[TrickyNekro]

I think one of the biggest problems with creating the robots we want is that we don't have a good means to power them.  Batteries?  Too heavy and not powerful enough (without spending a fortune). 

Well that's the 21th century problem, not just for the robots. And nuclear power is dying (the fission one, the fusion is to come).

And don't think that thermonuclear generators produce so much power. Plus they are still radio-active.
And if you don't give a darn for your little friend cancer, electronics do. Radioactivity destroys transistors.

And you wouldn't want some kilos of lead just sitting around.

Fusion is the way to go nuclear today... Not that it doesn't produce radioactive waste, but there are much much less harmful. (or should I say active)


In the end... Well... really? Why Skynet... I see no heavy reason... But well sick minds exist...
Still... they need to solve the energy problem first... then we'll see... ;-)

[Admin]

One major improvement in robotics we have seen is the price. Everything is orders of magnitude cheaper than it was just two decades ago, and the prices keep dropping fast. When I first started robotics 10 years ago, a quality dual motor driver cost $250 (in 2001 money). That same driver today only costs $50 (in 2011 money), yet it's smaller and better at the same time. Computing processing power/memory goes without saying.

That said, as Gertlex hints towards, robots still cost more than cheap immigrant labor

At some point robots will become more cost efficient than humans, given those trends, and that's when you can count on losing your job having robots doing your yard work/dishes.

[Killerwolfbot]

I think the main thing i would like to see is humanoid fighting robots. That would be so cool but cost alot in repaids and such.   Has anyone ever tried to build a tall humanoid robot? I know i asked this already but his is someting i plan on atmpting and would like some insigh.

[waltr]

A place to see how far robot have come is in some of the robot competitions. Rob-soccer is one of these that now have two classes of bi-ped robots. Google this for some great Utube videos.

[corrado33]

At some point robots will become more cost efficient than humans, given those trends, and that's when you can count on losing your job having robots doing your yard work/dishes.

I work in paint (currently) and one of the hardest thing we do is color match.  It's harder than you think to match a wet paint to a dry panel/color.  (You'd never imagine how many shades of white and tan there are...  ) 

But, there is already a robot to the same job.  It adds some tint, makes a panel, puts in in the oven, takes it out of the oven, the reads the color.  Then it does it all over again.  It's pretty simple actually.  However, it's probably very expensive and we don't use it yet. 

[mr roboto]

[youtube]Who's Newton? SynPet VHS Promo, 1989[/youtube]
this was just 20 years ago
cost about 8 grand
that was super expensive back then
now you tell me

[mr roboto]

oops
bad link
Who's Newton? SynPet VHS Promo, 1989