The current state of robotic's technology and actual creations

iamdenteddisk

This was an idea that came to me about why the current state of the art,cutting edge,top of the line robotic creations ,for lack of better terminology "suck buttermilk" .

·I spent my last three days combing through hundreds of websites,communicating with friends who are currently working as robotisist in industry,commerce and military jobs who all agree,we are meant for so much more than rebuilding the same kits that where designed decades ago for prototyping and discovery purposes, dont get me wrong .Especialy ameture's and novice beginer's should start there but why is there no funding for delving further into the realm of terminator,towering 30 foot tall mech's and movie type transformers?
why do we as engineer's continue to be hampered·into rebuilding what might as well be automated skateboards whit only a sensor or two to change in the last decade.·this should have evolved into much·greater creations since the 70's . As with all things money is the issue though current technollogy permit's these greater things the cost of building research&development is under some sort of control wich is not admitted to by any governing factor .I looked into building my own monstrosity even with all R&D done pro-bono still I cant seem to find two enableing factor's, a investor "for fear of not getting to proffit"·or a realistic insurance·coverage"this gives no good explanation as to its high unattainable price"·and this seems illogical to me to the point of turning my efforts to a different medium like video game programeing wich seems to move at such a pace you have to be a major corperation just to move with competitive speed.

realisticly go interview the builders at the contest trade shows and build off competition's what you will get is this
builder 1 has an automated skateboard that can strike a match.
builder 2 has an automated skateboard·with an fire extinguisher.
builder 3 has an automated skateboard·that works a rubix cube.
builder 4 has an automated skateboard that bounces off the wall untill it finds a door.
builder 5 has an automated skateboard that hurls soda cans.
Im sure you get my point there is nothing wrong with these so called competitions but these fact's

just today Vince Mcmahann sits in a comfy office ,probably behind·a grannit desk with ivory trim for pitting two humans hyped to bursting in a ring.
·Is there no business man out there with the power,gutts,money enough to get the same started with "mech wars" ?
he could rest assured that even total destruction of robots there will still be useable parts.
Also the current limitations of mechwars only allowing R/C opperated bots ,clearly an attempt at safty that could··be lifted because since the perfection of monster trucks we have had remote kill switch's so whynot allow full automation and use these kill switches, compleate enclosure is still usefull to stop flying debri,·but if we had the above then a group of robot makers could work in teams to bring creations to an arena for pit match's· get a lower insurance rate and we are good to go,and perhaps its like erco's signiture "build it they will come", just maybe if a group of builders did build it -investors,insurance company's aswell as spectator's will do just that.
··

Mike Green

There are some very interesting things being done in university settings, particularly in Japan on humanoid robots. There's some excellent research and development work on complex realistic facial movements and on walking, again all done in university environments. It is difficult as you've noticed to get funding. There's very little available from the US government except related to military applications. There's some funding from foundations and from large corporations (associated with very large "strings"). That's the way things work. In today's tight financial environment, you are unlikely to find any significant source of funding for small scale projects although many chip manufacturers are very helpful with samples if you're not too greedy or needy.

Specifically regarding large independent robots / terminators / etc. They're potentially dangerous. Even something like Parallax's Propeller QuadRover can easily hurt people and their property. That doesn't say that they shouldn't be built or used, but careful thought has to be put into safety and the protection of people and property. Insurance companies are sometimes overly cautious, particularly with new or unfamiliar things or situations. They are sometimes hard to convince and they tend to be more emotional and irrational rather than thoughtful. On the other hand, you're asking them to take the risks and they hate to lose money.

pharseid

· There's some good work being done with dynamically balanced legged robots. I think there's potential for some real fun there (anthropomorphic robots). Part of the problem may just be laziness. Legged Robots That Balance was written, like, 20 years ago and the typical desktop today has at least an order of magnitude more computational power than was required back then, but how many people have even attempted to reproduce their work? I've thought of using a tethered partially supported humanoid robot to try to work out the issues of a dynamically balanced robot, but it's an awful lot of work to develop software for a robot I couldn't build anyway. If you're looking for·a good project to work on, there you go.

-phar

Al2

People are not watching destruction of robots as much anymore, so why would someone want to put their money into it? There is way too much destruction of things anyway. Now making robots that can help you, I am all for.

http://robotics.scienceontheweb.net

MarkS

I think many people on the bleeding edge of robotics tend to over think things. Not that the future of robotics is easy, but certain problems are not as mountainous as they would like to think. The human mind, for instance. When most people try to create a facet of the human mind, they try for the end product. However, this ignores the way the brain develops. The human brain is almost 100% hardware, with the individual connections making up the software. These connections form while learning. We have given ourselves an interesting platform with our current computer technology. The basic structure is present with the hardware, all is needed is software to get something remotely intelligent. The software would need to start small and learn. Take the sense of hearing and the ability to speak for instance. A baby coos and makes noises, not to amuse its parents, but to form the structures in the brain necessary for hearing and speech. As the structures develop, the sound patterns start to become more coherent. Of course, there are many layers to this, vocabulary for one, that would need to be addressed, but rather than trying for Star Trek's computer's voice recognition and speech abilities on the first shot, we need to try for a baby. Properly written, and I still haven't the foggiest idea how, it wouldn't take years for a computer to learn hear and speak since the structure (hardware) is already present and fully formed. This is, of course, over simplifying things more that a tad, but our robotic technologies seem to be stalled because we keep trying to leap past all of the lengthy and boring stuff and get an end product to market as quickly as possible.

CHIPKEN

The problem with robotics today is the mechanical·components and powerful batteries are not there yet. I guess for the first time is history the electronics available·exceed the mechnical hardware available. This was not the case for a long time.

Chuck

MarkS

CHIPKEN said...
The problem with robotics today is the mechanical components and powerful batteries are not there yet. I guess for the first time is history the electronics available exceed the mechnical hardware available. This was not the case for a long time.

Chuck

Yup. There is really no technology that can compete with biological muscle in terms of power to weight ratios and current consumption. Air muscles and electro-active polymers (EAP) can actually exceed a biological muscle's power to weight ratio, but there are severe drawbacks to each technology. Air muscles require a compressor and proportional valves both of which are bulky and power hungry and EAP's require kilovolts of power to activate, albeit without much current draw.

CHIPKEN

The proposal that John McCain has for a $300 million development effort for a better battery is probably a very good idea. For right now we will have to use·a hydrocarbon fuel·power, like the Quadrover has, to overcome the lack of a better power driver. At least this is a more advance approach than what MIT and the others are using with their battery powered robotic units. My suggestion is to build robots that do something and hope that the electrical power packs will follow in the future.

Chuck

pharseid

·To MarkS,

····· There are also electrochemical polymers that are sometimes lumped in with EAP's. They use quite low voltages (single digits) but are slower than human muscle. But they would still be interesting if they were commercially available.

······As to AI, there has been a lot of work done on self-evolving behavior. I would guess that the researchers in that field have a pretty good idea of how difficult the problems are. Years ago I had similar thoughts to yours on evolving artificial intelligence, but in my casual research of the field over the intervening time, no obvious approach has ever presented itself.

·-phar

iamdenteddisk

I started this thread in hopes to trigger a "group effort" to I guess allow advancement of technollogy and also I should say public technology awareness because in actual truth we can now build machines with their own power source on board eliminateing the fear of useing up the bateries to quickly, maybe Im speaking latin to some of you but let me give some examples here.

as to power availability these are some things that currently work on my projects and support info can be found online.

1 solar engine to drive a dynomo charger ,this allows continous use in lighted enviroments the use of the solar engine instead of direct solar chargeing is its current production ability.

2 a stirling engine to drive dynomo chargeing this allows thermal energy to be used like the solar engine above and these can also be insulated well enough to not only use heat but also cold tempurature to drive dynomo.

as to artificial inteligence I started playing with this in some forms it shows some promise for speech engines but I will later explain better methods

example a

the machine is asked a question"IS THIS MY CAR" the needed information is encapsulated in the question to answer if the question evaluates true"THIS IS MY CAR" notice this is the only possible legable sentence that can be made from the provided information ,only a few modifications to be made -MY should be replaced by YOUR as it is possesive and must reversed in response and also the secondary object in the sentence must be further described as in "CAR" machine response 1 "define your CAR" ?machine response 2 "define this CAR"-if both quarry's compare and evaluate·same or true then machine response 3 is "THIS IS YOUR CAR"

example b

now the better method -anyone who has had calculus should recognize this-

Esummand/k(n,k)=answer(n)

or

E/k(n/k)^2=(2n/n)



where the left of equal sign is the question and the right is the answer respectivly,and the answer must be made possible either in a list of actions or responses.

the capitol E was used to represent summond wich was not available in font.

or simplified for those useing only the paralax bords for logic

let 10 forward steps represent freedom to move and all else be considdered function when a whisker is bumped try each function untill freedom is had this is a little cheesy but it simply works.

·also for those who want to see deeper into the subject of machine inteligence /not so artificial but calculated I also tried to·attach some must reads gathered in my reserch that wouldnt upload,so any one wanting to further research may contact me. iamdenteddisk@yahoo.com

I have also a unification/"Everything" theory that Im trying to expose to the intelectual comunity,·I sent email to steven hawking & some others·but the only response I have gotten thus far is confermation of delivery. any who want to see may send request to my email only those with U.S.·mail·adresses are responded to ,sorry for that but we are·currently envolved in a·war and a treason charge isnt what I want·.