"Screwdrive"

[NVR2FST]

I believe robotics is growing in the US. Personally I think Robobasic is a turd . I imagine it is fine for a programmer but not a more casual or beginner to robotics. I know that sometime in the relatively near future as robotics grow stronger in popularity for the average person (price and ease of use) we will see much more intuitive software. It is natural progression. From DOS to Windows for example. Before you had to do all the work and Windows came about and did all the bullwork in the background with minimal input. The same will be with Robotic software. We will look back and say, "man that was silly back when". We had to program everything and we were paying $1000 dollars or more for a "basic" robot!! I have had my Robonova for almost a month now and I am ready for something more "modern" or up to date. Dont get me wrong I am not putting it aside, I just want something better for the money. If you think about it, these robots are still at an introduction level. If they started years earlier and made there way into the mainstream earlier, we would be seeing some really amazing robots now. We have the technology. It just needs to be implemented into the mainstream even more. Thats where funding will come from and progress robotics in leaps and bounds. Look at our computing power we have now. It is just amazing. They just need to get the software rolling. I have seen VERY basic Robotics and more complex ones, but nobody has hit the sweet spot yet for the mainstream . They are way to expensive for what they can do. Your average guy wants more. Only because he knows it is possible by looking around him in the world today. I see all of these people cramming wires, boards, sensors, gyros, crude modifications, into and onto these robots that almost make them unusable because they dont have the room or a well designed platform which to place them. I believe that at this stage in robotics they are making the platforms a little too small. That will come in time. As soon as somebody comes up with a properly sized robot with a well designed frame, good software, and a reasonable price. The mainstream will bite. Some company needs to take the initial "hit" in development and they will own the market. Put away your flamethrowers, I understand all of you people with your countless hours of building, testing, and modding, have kept this going and excelling. Hell, I have been doing it for as long as I can remember with other things one being my RC helicopters. Have you guys noticed in the last five years how the helicopters have progressed! Before you had to "fly the tail" which kept it from the mainstream. Now they will almost hover themselves. Some of you may say, some people just dont want to have to do any work with there hobby. Not in my case, I would mod the space shuttle if it was mine. That would never stop. I just believe we are behind in what is definately posibble today. I have many ideas on building a biped robot. I have a mill, lathe and many other tools of creation. The one thing I would like to do is use "Srews" to move the limbs. I mean something similar to a worm gear. The "Screw" would thread itself into a threaded tube that was connected to the opposing limb. Powered by a servo motor "Micro step motor". This would be very fluid, full of torque and rigid. I dont like the idea of using servos as the frame. It is a huge waste of energy even at idle, just holding itself up. Servos buzzing constantly, wearing themselves out, wearing down the battery. And in the end it isnt very rigid. The robot still will "Flex" the servo drivetrain. Antiproductive if you ask me. If you create a "skeleton" first and then add servo motors to it , it would be much more solid feeling. My only problem is making the "srewtubes" as I wouldnt want to use standard machine screws. It would be more like a worm thread but fine. I believe the female tube would be a slick durable plastic thread. Brass is a possibility for a home build. If I could find something premanufactured like this it would be great. Maybe something out of a cd player or something with an ejection drivetrain. I believe I have seen something like this. Not the gear type. The idea is too have no flex and to stay put at idle or any stop position. A worm type setup cannot flex or be pulled. It would look similiar to a ram on a peice of heavy equipment but obviously not hydraulic, but with a screw. This look would also enhance the appearance of the robot. It would look very "tough". The "Screwrams" could be quite small. They could fit inside or outside the robot frame and look good either way. BURY THE WIRES! This is one thing that blew my mind about Robonova and these other robots. If any of you have any ideas on this, let me know. I can see it in my head, I just need these "Screws". I got the idea from when I was a little kid. I used to Mod my toys from a young age. One of them was a Tonka dump truck I had. I mounted a long machine screw on a small electric motor shaft. I then mounted it to a pivot on the truck frame. I put a nut on the dump bed with a slide mechanism. Although it was crude, it worked perfectly!! From that day on I was hooked. I ran to show it to any one who would listen including my friends and my father. My father was quite impressed, considering my age. Some of my friends liked it too. But there are always "Them". Those people that just dont get it. "Why do you waste your time with this stuff?" "You have too much time on your hands". Thats when I learned I was "Different". Not that I was antisocial or anything. Far from it. I enjoyed building things as much as I did partying with friends. Now I am older and "They" are still around and asking why I waste my time with this stuff. I just say "As opposed to doing what.....drinking all of my free time away". They usually shut up. Ironically, who do you think is the goto guy when they have a problem? Some people accept drinking and partying as the multinational passtime. Not me. I enjoy creating things. Learning how things work, moddiing etc. Dont get me wrong ,in the summer I will "throw a few back" but I will always enjoy the satisfaction and the sense of accomplishment that these "Hobbies" give. So friends, dont let anybody make you believe what you are doing is a waste of time (Sometimes a waste of money) but never a waste of time. Does any of this sound familiar? Somehow this turned into a blog, I think Oh, once again if anybody has any suggestions about the "Screw", please let me know. I have no doubt that this would be an excellent design if I could get this one thing figured out. The brackets and all would be relatively easy (time consuming) but relatively easy.

[NovaOne]

Interesting ideas

Correct me if i'm wrong, but your "screwrams" sound like (electro-mechanical) linear actuators Bren is using some on the ankles of his home build.

http://en.wikipedia.org/wiki/Linear_actuator

[MYKL]

I've built some screw based actuators and they work fine. Pulling weight is more smooth than pushing for some reason. I think a solution in this case is to place the electronics for the servomotors on a seperate board so that the space taken by the mechanical parts is kept to a minimum. There is a company that makes a very efficient screw drive system. It has a planetary set up that removes backlash from the equasion. They do not (as far as I know) make a set small enough for the Hobbybots yet. I'll look them up again and post a link here for you.

I've been working on a liner acutator design too. I'll get back to it after my hands are done. It's an servo driven Hydraulic muscle. I think you'd like it...

Keep up the Imagine based design. Its the only thing that will allow us to hope to compete/play with the Japanese.

[NVR2FST]

Thats exactly the idea I am talking about! Now to just find the right size as they would need to be small. I am going to run with this and see what I find.

[Robo1]

Have a look at firgelli for the linear activators. I use the PQ-12 in my biped.

Bren

[MYKL]

Here is an example of a PLANETARY ROLLER SCREW:

http://www.nookindustries.com/RollerScr ... ossary.cfm

Those Firgelli' look cool but seem a bit slow. Do they take a lot of power?

[NVR2FST]

Dammit, you guys are bumming me out!! I thought I was on to a new idea! This idea has got to be way better than the servos ! Can I see any in action. Why in the hell arent these "THE WAY"?

[NVR2FST]

Now to tell the wife I need to spend another couple thousand!

[NovaOne]

I may be stating the obvious but, catch and play is quite an important programming method at this "early stage". These types of linear actuator would prevent you from using this useful technique.

[NVR2FST]

You are right, yes I did overlook that. With the rigidity you would lose the ability to manipulate the robot physically. But as dumb as I am with code I have been doing quite well with Roboscript and moving the sliders to a position and saving. Then repeat. I actually prefer this method. Is there any way to copy a line that I created and place it later down in the code in roboscript? With even a basic program this would appear to be a no brainer to implement into a program. I mean it is text. I should be able to right click a line I created hit copy, move on down and paste into a new line. These little things just make me feel that this program is below average.

[MYKL]

I'd like to see a GUI that works with keyframing like animation software uses.

You build a skeleton and set servos/actuators at the joints in a virtual context that reflects the Kinematic constraints of your 'bot and animate it in the machine while it uses a physics engine to add gravity and motion 'effects' to what you animate. Then after you have your virtual bot doing your bidding in cyberspace you have the program kick out a bit of code for your board and watch your bot to see where you need to tweak for finished movements.

Realtime control is where I wanna go. You wear a haptic interface that send signals to your machine to flail about in a perfect cloned performance of what ever you are doing at the moment. MAN-MACHINE INTERFACE baby...

You could record sets of movements and use them to program algorythems for mundainities such as walking on a flat surface for a prescribed amount of time.

I want to build a suit like the Landmate in the Appleseed Anime.

[MYKL]

Have you seen FESTOs air muscles?

[Robo1]

Hi All

I may be stating the obvious but, catch and play is quite an important programming method at this "early stage". These types of linear actuator would prevent you from using this useful technique. Confused Question

On the firgelli there's a pot that you could use for this. You can move the shaft but small amount 1%. So if you put this in a control loop that looks for this and move's it in the force direction. So it can be done, don't as easy but do able, should be doing this soon but have a load of other thing that have priority.