Operational parameters for industrial biped.

Given that industrial biped will be common place within the next couple of decades, does anyone have any suggestions as to the specifications for such
devices ? There appears to be little or no research on the subject to date ( I guess Aliens offered some suggestions). In particular, I would welcome comment on the following parameters ;
1 ) Acceptable start up time (it's not like you would just walk up and turn a key)? 2 ) Useful payload? 3 ) Duration (how long should it run without fuel or charge)? 4) Level of expertise of operators (should it be idiot proof or is a level of technical ability required)? 5) Cost (what should you expect to pay for one)? 6) What would you use it for ?
Thanks,
Joe Cronin
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
Why would a biped offer the best functionality? Because we humans only have two legs? Far more creatures on this earth have 4, 6, and even 8 -- a crockroach is at the top of its evolutionary ladder, and has existed for several million years, so the design must be doing something right. Or consider: a six-legged and even four-legged creature can survive with a missing leg. A two-legged creature will die.
I can see the validity of a legged machine to traverse a wide variety of terrain, but I don't buy into the notion that the ideal industrial robot will be a biped. Could you provide more details on why you've narrowed the configuration to just two legs?
-- Gordon Author: Constructing Robot Bases, Robot Builder's Sourcebook, Robot Builder's Bonanza
Joe Cronin wrote:

Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
Gordon McComb wrote:

------------------ Nawh, it has six legs just because it was designed before life had sufficient processing power to actively balance.

---------------------- Not a robot, it will hop.

--------------------- Lighter, simpler, more flexible, takes less volume. -Steve

Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
R. Steve Walz wrote:

It was designed?!?! REALLY?!
Bigger doesn't always mean better. I don't think a roach has ever had to worry about injuring itself falling down/running into something. Or having a battery fail, or a circuit board short out because theres steel particles in it's local atmosphere.

Really? will it hop up stairs too? Will this functionality be available in the 10-years-from-now-model, or will it have it before?

lighter? An industrial biped robot, lighter? You think they're gonna make this thing outta balsa wood or something? The motors ALONE are going to put you in the 100lb range (this is industrial, remeber!). Add power supply to that, control circuitry, sensors. Add to THAT a structural body that can support the weight of the robot itself, AND a load...Light my ass!
SIMPLER?! Than what? A starship?
More flexible ... thats true. So long as there are some decent advances in power supply systems (Micro fusion power cells, perhaps?). And some nice leaps in structural metal technology. Of course, you can't really circumvent the laws that govern our universe. I don't think I would want a robot+load with a 600lb gross weight running by me.
Less volume? perhaps.
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload

I don't accept this as a given. Most industrial machines today use wheels or catepillar treads. Wheels are simpler to design and control, more energy efficient, permit faster travel, less shock and vibration, etc. There are some environments where legs work better, but I would not describe those environments as "industrial". I don't see any reason to stop using wheels just because we are going from human operation to autonomous operation.
Even where legs are useful, four are usually better than two, for both speed and stability. Bipedalism seems to me to be an accident of evolution. Evolution works by incrementalism, and each incremental change has to be an improvement. So you can slowly add functionality as you change the front legs and paws into arms and hands, going from possum -> lemur -> monkey -> ape -> human. But there is no way an evolutionary process can just stick a pair of arms onto a quadreped, even though that probably would have been a better design. If centaur- like creatures actually existed, I think they would have wiped out humans long ago.

If it is gas/diesel powered, it should take a second or two to turn over the ignition. If is electric, it should be "instant on", like my PDA.

Why not?
2 ) Useful payload?
The most useful robots would be drop in replacements for humans, so they should be able handle the same range of payloads.
3 ) Duration (how long should it run without fuel or charge)?
It depends on how quickly they can refuel. Batteries take a while to recharge, but a good technique is to just swap out the battery and get right back to work.
4) Level of expertise of operators (should it be idiot proof or is a level
Operators? Robots should be fully autonomous. They should not require operators.
6) What would you use it for ?
What would be really slick, is if I could buy a robot that could type on a computer keyboard. Then it could write my software for me.
-bob
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
On Sat, Mar 06, 2004 at 04:06:07PM -0800, Bob wrote:

I think the typing part would be trivial compared getting a machine to understand your specifications and turn that into working software. This is even a non-trivial problem when working with human programmers, let alone, robot programmers. I guess if you could solve that problem, you will have achieved what researchers have been trying to do for decades - create a machine that understands human language and ideas.
-Brian
--
Brian Dean
http://www.bdmicro.com/
  Click to see the full signature.
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
Bob wrote:

--------------------------- Look up SegWay, and the latest balancing wheelchairs by the same company. While wheels seem important now, the new active balancing will quite shortly blow them all away.

-------------------- I think you missed the point. Without dynamic hopping, the simplest legged creature is bipedal. Not merely mammals, but birds of all sizes as well have reduced their weight by using only two legs to walk.
The next 5 to 10 years will bring us bipedal robots with better balance and recovery than WE have, consider the last 5 if you don't believe this, and they will be cheaper and lighter than wheels in 10 years.
They will be able to RUN up and down stairs and avoid us EVEN while carrying heavy loads. They are now working on legged "wheel"-chairs that will enable even paralyzed people to "walk" easily and safely.

----------------- We'll have bipedal robots with arms which have better balance than WE have LONG before they can do more than follow VERY simple verbal commands.
-Steve
--
-Steve Walz snipped-for-privacy@armory.com ftp://ftp.armory.com/pub/user/rstevew
Electronics Site!! 1000's of Files and Dirs!! With Schematics Galore!!
  Click to see the full signature.
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
Just my 2 cents -
A wheeled/tracked robot won't fall over (onto someone, or onto delicate machinary) if it experiences "technical difficulties". A wheeled/tracked system is not as electronically complex as your humanoid/bipedal "Robot worker of the Future", and thus has less points of failure.
Do you think a bipedal robot can be made 100% fail-proof? If you do, would you be willing to pay the doctor bills/repair costs if someone gets hurt (or killed...a humanoid robot wouldn't be terribly light), or if some multi-million dollar piece of equipment gets bashed?

Even the segway is under a great deal of scrutiny, last I heard...You might consider the possible reasons why...

The simplest legged creature is a biological marvel that we'd be hard-pressed to mimic, even in a hundred years time. Just because nature can do it, doesn't mean we can

Even with this "better balance and recovery" that you say we'll have in 5 years time...can you guarantee that they will never fall over?
A human can fall over all they want and not get people killed (except in a few rare cases). A 300lb biped robot, possibly carrying a significant load, is quite another matter.

LOL. Sure. If they're equipped with anti-inertia devices. I don't think theres any creature over 10lbs that can stop on a dime while running. You get anything over 100lbs moving at a running pace, and you're gonna be hard pressed to stop it, or have it avoid anything, ESPECIALLY when carrying heavy loads. Another thing to think about - It takes time for a human being not to flinch (at the least...duck and cover might be the best description for a first-time encounter) when a moderately-sized biped robot is bearing down on him/her. It will be a LONG while before we have systems that can do ANYTHING, while taking the silly humans into account. :)
Those legged "wheel"-chairs will probably be ready for public use after dozens and dozens of redesigns, several advances in power technology, and about 6 years of product testing, to ensure its actually safe enough to use (and there is always the possibility that it just WONT be safe enough).

Steve...I'm all for having faith in technology, but lets not take it THAT far, okay? :)
-Alex
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
R. Steve Walz wrote:

There have been balancing problems with the Segway. I'm still looking forward to seeing the wheelchairs being sold.
However, active balancing takes constant application of power. Being able to have a static balance position is a useful thing as long as our power supplies are limited.
While I would love to see an industrial strength bipedal robot, I doubt the 10-year timeframe. -- D. Jay Newman
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
Why should an industrial robot be a biped?

such
turn
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
Bipeds is only a "fad" at the moment. It seems everyone is on the "more human like" bandwagon at this time. It'll pass, machines should be built to perform a job or jobs, not look pretty. It's like trying to make a bulldozer into a pretty biped human like thing, once you do, it isn't a bulldozer anymore. Or why have a biped machine drive an automobile, build the robot into the car instead, gives you one more seat for passengers. Look into past history of animals and you'll see Quadrapeds have been the dominating mobility method for hundreds of millions of years on land. Evolution doesn't keep mistakes around very long. Although insects show hexapod and octapod mobility success. Large animals were all quadraped. Bipeds were always in the minority, primarily in predators, in the past. Bipedalism today only exists in birds and humans. Humans paid for the bipedalism with bad backs, leg joint, spine and pelvis problems too. Machines would have the same problem. But quadrapeds exist all over. Four legs is a big evolutionary advantage, you can still move if one leg is injured or fails. Although hexapods and octapods offer some advantages, we don't see large six or eight legged mammals or reptiles wandering around, nor have we found any on the fossil record either. So apparently hexapods and octapods were bad evolutionary ways to go for larger animals, but very effective for insects. Obviously then, one should look to quadrapeds as being the most efficient and effective way to achieve mobility in machines using some kind of legs. But wheels still offer the most efficient way to get mobility without using legs.
Anyway that's my opinion.

such
turn
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
On Sun, Mar 07, 2004 at 10:37:45AM -0600, Earl Bollinger wrote:

One good reason for "human-like" machines is that, if constructed well, they can be made to be general purpose. One machine could be applied to many problems instead of having to make many different kinds of special purpose machines. Human-like machines could use existing tools made for humans so you could go down to your local Home Depot and buy a DeWalt cordless drill that your "human-like" robot could use for $200 instead of having to buy the $8,995.99 custom drill attachment from the robot manufacturer. I could see a lot of advantages to that. Not only could it drive a vehicle, but it could use any tool or device that is designed for a human, as well as go into buildings made for humans, etc, etc, etc. In a world made for humans, there's a lot of utility to making a machine that fits the world instead of reshaping the world to fit the machine.
Of course, I have yet to see any mechanical thing come even remotely close to the utility of the human hand. But if it were possible, and affordable, it would open up a much larger problem space for which robots could be applied.
On a related note, I seem to recall that about 6 or 8 months ago there was a guy named Bruce, I think, posting in this group talking about his new "synthetic muscles" or something like that. He was trying to form a team to make a humanoid robot using the stuff. He was saying he had solved the critical problems, and was now trying to recruit people to help out with the rest of the design. Does anyone know if anything come of that? I haven't exactly been holding my breath waiting for the announcement, but a little follow-up would have been nice, especially after the hype. Maybe there was, and I just missed it.
-Brian -- Brian Dean http://www.bdmicro.com/
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload

Polytechforum.com is a website by engineers for engineers. It is not affiliated with any of manufacturers or vendors discussed here. All logos and trade names are the property of their respective owners.