bug

Let's Make Robots Robots
1

UPDATE I stopped working on this version, the new version is here https://www.robotshop.com/letsmakerobots/bug-20

I have created my first part of my first robot, I am ready to make my entrance in this section of LMR. My robot will be modular, I will start with 3 modules each with 2 legs. This will classify the robot as an insect. If I manage to create more segments it will be a centipede I guess.

If all works well I will submit my robot to the one motor chalenge.

The servo will rotoate the main axis from -80 to +80 degrees for straight movement. From 0 degrees to -80 and back, the odd legs will go up and down. From 0 degrees to +80 and back the even legs will move up and down. So always half of the legs will be down, which means no loss of energy.

The forward - backward movement is a little tricky because it is out of phase with the servo movement, it lags behind 1/4 of the cycle. I designed a mass-spring system for each segment (the mint-green parts). Empirically I will have to tweak it so its frequency will match the oscilation of the servo (between 0.5 and 1 herz). This will have the effect (I hope) that the insect will start by just moving its legs up and down, and then slowly gains speed. In theory a lot of the kinetic energy is preserved, so the robot should run quite fast.

Steering is the other challenge. Each segment adjusts the leverage of the forward-backward movement of its legs depending on the postion of its front neigbor. So basically the thing goes where the head goes. I can use the remaning 10 degrees of the servo to move the head, using the same servo. So when the servo moves from -90 to + 80 the centipede will go left.

I plan to make a small series of this robot. Anyone interested in getting a free prototype, in exchange for feedback? The prototype will be laser-cut from polystyrene plate.

These are the parts needed for each segment

build day 2

today I built the forward leg movement. Here the parts:

And the working demo is in the second video

update sept 15 2014: one leg is moving: see 3rd video

update sept 16 2014: and now it crawls without making progress! see video 4.

 

This is a companion discussion topic for the original entry at https://community.robotshop.com/robots/show/bug
2

Great work…

Nice work and it looks great. I dont really understand the mass spring system. I saw your video on other post. Anyhoo, may i speculate? It seems to me that it creates a torsional force. To create a linear movement, should it move like the opposite oars of a boat? I mean -in the same direction. Correct me if i am wrong. I am no expert haha.

3

I dont think it works as you say

First, I have seen your video and I am impressed. I am also glad that we didn’t coincidentally came up with the same solution :).

To answer your question, I don’t fully understand how you think my mechanism works, but I don’t think it works by creating a torsional force…

I am not sure I can explain this clearly, in text it sounds very complicated. For those who don’t understand, I will make a video once it is working… But here it is, the explanation of the technical challenge I have. Read slowly and try to visualize it:

Divide the servo oscillation in 4 steps: A) from 0 to -90 degrees, B) from -90 to 0, C) from 0 to 90, D) from 90 to 0. Then the vertical movement of the legs is this: A) odd up B) odd down C) even up D) even down. Note that if you make A) work, B) will automatically work too, because it is the same movent but reversed. Same for C) and D).

In other words,

Now for the horizontal movement you want during A) and B) the even legs (the legs that touch the ground) move backward and the odd legs (the legs in the air) move forward. And the roles are switched during C) and D).What you DON’t want that the even legs move backwards during A), then reverse during B). It would let the robot move back and forth, without making progression.

The problem is that during A) and B) the servo changes direction, but the legs should move in the same direction. And the same for C) and D).

The way I want to solve this is a resonator that follows the movement of the servo, but with a delay.

I made a test rig, and this is the video you refer to (https://www.youtube.com/watch?v=pu9EAU266io). The problem with this video is that the frequency of this rig is much too high. Imagine the same mechanism but with a very weak spring, so it will resonate at 1 cycle per 2 seconds (veeeery slow). Then you will see that the arm with the weights will move out of phase with my finger. In other words, it lags behind. I will upload a new video at the correct frequency once I have received the materials I plan to use.

What I hope is that I can make this phase shift 1/4th of the cycle. So when the servo does ABCD, the arm will move DABC.

Okay, it sounds like crazy. And maybe it does not work at all. We soon will know.

4

Arrrghhh!
I have a similar idea for a turtle robot for another one-motor challenge. The servo would wag back and forth to move the legs. If the servo went a bit further to the right or left it would turn the turtle a bit to that direction. Completely right would shift gears to cause it to back up and completely left would shift it back into forward.

I hadn’t thought about the resonator, but I’m not sure it would work on a turtle anyway (four legs, crawls rather than walks).

5

Why arrghhh?

Why arrghhh? Just build and submit it. What do you mean “another one-motor challenge”, is there more than one?

Good luck.

 

6

I don’t have the time to
I don’t have the time to build a second robot yet, nor do I have my 3d printer (Real Soon Now). Right now I’m working on an inverted pendulum bot, programming a Teensy 3.1 to be a generic controller of such things as well as the body.

Eventually there will be a second one motor challenge or I’ll start one.

7

**That read badly.

I mean I’m**
That read badly.

I mean I’m programming the Teensy 3.1 to be a generic balancing robot controller, and I’m building the body of the test robot.

8

thanks for taking your time to explain…

your explanation is very good and vivid. I am sure it is going to be  one cool bot…