Attempting to build an ATAT

I don’t remember in the Movie how they did it, but if the Total Motion was modeled after Elephants, not just the Leg Motion, then it would be “Left Front/Right Rear then Right Front/Left Rear”.

The Center of Gravity is the Key, how ever you get it to work.

Thanks! I’ll do a little testing, see if I can get the motors synched up for the alternating gait.

Edit: that is, if the feet have enough traction to do anything besides slide randomly…

So I haven’t been able to get the gait synched up properly, mainly because the mentor I’m working with for programming asked me to remove the ‘training wheels’ for testing. So we still really can’t get it to balance. I’m thinking of changing the default standing so that the legs are straighter. I know this’ll decrease the size of the support plane, but that may also counteract the falling-over-ness of the robot. I’m not sure, but I guess it’s worth a try.

Also, I put limit switches on the feet to help the robot know when its feet are on the ground. Right now my mentor and I are working on coding up some potentiometers, and the little thing hanging off the right hand side, between the hip and the knee is the one we’ve got plugged in. It’s not mechanically mounted at all, mainly because since it’s not a VEX pot, it’s a little hard to integrate the gearing and whatnot. So any suggestions for connecting the pot shaft to the VEX gears is highly welcome.

If a shaft collar fits on to the end of the potentiometer, you could glue two shaft collars together, stick one on the potentiometer and one on a shaft.


We actually tried drilling out the center of a 60-tooth gear to the exact diameter of the pot shaft, I think it was 11/64? Turns out, the fit was perfect and there’s enough friction that it’s practically impossible to turn just the gear and not the shaft. So that solves that.

Now all I have to do is get enough pots and then attach them to the robot. crosses fingers

Cool! Good luck!

So all the potentiometers are mounted… Those are the little goldish disks. Now I’m working on programming in something my mentor refers to as a PID (proportional integral differentiation, I think).

I have success!

So it doesn’t quite walk yet, and there were a few glitches. But here are the two videos I took, as well as a picture.

This one was the original test to see if the motor could center the potentiometer properly.

This one is me testing out how strong the motors are, like how much weight they can resist. Watch for more details. My mentor helps with taking the video.

And the attached picture is a photo of when we got it to balance perfectly on three legs. My mentor’s hand is under the foot, but he’s not holding it up!

I don’t get why everyone in Star Wars seems to make all the droids (except R2D2) with legs and be a humanoid. It’s so inefficient, costly, and just stupid for most situations. And it’s slower than a robot with wheels. But, I guess it is kinda cool. Great work with the robot though. Will it only move one leg at a time so it can stay balanced?

Well, the main reason I chose a walking robot is for the challenge. While something that rolls is more efficient in many situations, it’s quite simple to build a wheeled robot. Walking, however, is one of the hardest things to get a robot of this size and type to do, from what I’ve seen in my research.

As for walking, that’s still in the works. I’m not really fluent in code so my mentor is teaching me as we go. It turns out that EasyC might not be the best fit for this job, so we’re going to try something else that will probably be a little more flexible. The gait is probably going to either be a left-front right-rear right-front left-rear pattern, one foot at a time, or a left-front-right-rear right-front-left-rear with two feet at a time. I’m not sure which will work better, but the ideal situation would be for it to walk like it does in the movies, with a left-front left-rear right-front right-rear gait. fingers crossed

The potentiometers are helping immenseley, thanks to Rick TYler. With those working, I think that it should be simpler to get the robot walking properly quite soon, depending on my schedule.