Well, based on the previous email thread regarding why there were no 3-legged bots out there, I’ve come up with this idea. It’s a combination of the Lexan-based Scout kit (top and one leg), and an SES humanoid biped (front leg design).
It’s not complete, as there are still issues being worked out, namely:
I need to design some feet that will support the weight based on specific stances.
The two front legs need to be farther apart, as they get in the way of the rear leg and foot. I’ll probably just modify the top two C brackets with new holes to offset the universal brackets.
Obvious need of wire ties.
Stronger servos - most of the ones I’m using are hopelessly inadequate.
He balances and can take a few steps using Sequencer. I think that this configuration works a little bit better, giving a longer step and balancing the CoG better. It also allow him to do this:
I’m going to try and design a 3-fingered hand that can spread wide enough to function as a 3-toed foot as well. If anyone knows of one, PLEASE let me know!
Hey, nice design and all but I don’t believe this could be said to be a triped. In a triped, the legs should be each 120d apart. You basically sticked a leg onto a biped. The biped is able to keep stability with its two legs down while the “third” leg is waving around. So I don’t qualify this to be a triped.
I’m not sure I follow your logic. If you look down from the top of the robot, each leg is, in fact, at 120d. Do you mean that each leg must face 120d? That’s possible, but I don’t know what rule is in place that says this orientation defines a triped. Also, when walking, all three legs are used. The triped is designed to be able to stand/balance on any combination of 2 legs. A biped can raise one leg and “wave it around” - does that mean it’s really only a monoped? A dog (quadraped) can raise one or two of it’s paws and still walk (and don’t face 90d) - does this mean dogs are not quadrapeds?
In any event, I actually find this triped design to have have considerable utility. I’ve got some experimentation ideas in mind that I’ll be exploring once I receive some new servos. Specifically, I think that the design will allow the robot to climb up onto, and over, an obstacle much easier than a biped can. It will be interesting to see in which direction this is most efficient - single-leg foot up and push from reverse-knee-legs, or one reverse-knee-leg foot up, pushing with single-leg (there is no real forward orientation on this robot, either single-leg leading, or reverse-knee-leg leading is valid).
I think the idea is that a tripod walker would basically be half of a round hexapod. If you remove 3 legs from a hexapod, you have a tripod. I think you might be able to make that setup work.
However. as long as it has 3 legs and they are 120 degrees apart, I don’t see what other qualification would be needed to call it a tripod. I see why the round hexapods are so popular, but I still want to build an inline hexapod/octapod.
I agree. I believe that’s how the tripods in War of the Worlds walked. They always had two legs down if I remember correctly. I think the third leg is more for stability when standing/positioning, rather than actually being needed to walk, but it does provide for an easier turn - just rotate the body and change the pair of legs it walks on.
I think that would depend on whether the biped has arms or not and can use them to aid in climbing. I don’t think I have seen a climbing biped robot yet That brings up a whole new question… We have two legs and two arms which we could use for walking… I guess it is just how many legs an entity has for classification purposes.
I’m interested in your progress here. I’ll be looking forward to videos as you progress.
hmm, just got an Idea that I would LOVE to do!!! Imagine a robot with the 3 legs as Johnny showed, but know with the torso of the biped Nick… As soon as I have eouch servos I will try to make the best 3 legged robot and then ad a body, and make sure the legs don<t come in the way so that it would be hable to climg on stuff!!! it would be the perfect robot!
or would you try to get equal steps out of each leg? LCRLCRLCR ?
or LRLRLRLR and C only when stationary? or interupted?
I think right now it is L+R,C,L+R, C,L+R,C ?
… quite reminds me of the octapod at U of calgary? where they try to make it learn to move and after each success they remove a leg and then get it to learn to move again and again remove a leg …
From my initial tests and observations, it appears that the LCRCLCR gait would be best for normal locomotion. I tried this myself on the floor, using my arms put together to represent the single leg in front. It would look something like this:
This is true, just as it is for hexapods with two basic configurations. There are other setups of legs that might work also. For instance, on my Walk 'N Roll bot, I am planning two sets of four legs, front and rear sets. I think it is all in where and how you place the legs.
Most people just think of the standard 120 degree apart configuration when they hear “tripod” used.
