Let's Make Robots!

SQ01 (simple quadroped 01)

walk

well this is what I've been working on for a while.

The plans started out as a hexapod with the same basic idea, but the servos I ordered for it are a bit weak so I had to do something with the three larger servos I had. The result is this little guy. 

The body and legs are made from clear acrylic. I't light and fairly easy to work with, and I've found a reliable and relatively cheap source for it.

It's held together with double-sided tape and a bit of glue. Only place I had to resort to more drastic measures was the middle servo that does the tilting.

I'm hoping to make the gait more stable and fluid with some tinkering in the code. shorter steps and less tilt might help. I also have to work out how to make it turn :)

later I might add sensors to make it navigate by itself, but I still want to work out some stuff about the movements. I'm not sure visual or ultrasonic navigation will work properly on this robot right now because of all the movement and tilting, so I'll have to work that out first.

 DSC00464.jpgDSC00465.jpg

 

 

Update 090103:

I've worked out how to make it turn now!

I use the fact that the robot hardly moves when the legs move independently. I make it rotate by moving the legs in sync and then reset the legs one pair at a time. it completes a 90 degree turn in three sycles, but I think I can improove on it some more.

it's the most efficient way I've found so far, but I'm going to keep at it. I'll hopefully have a video up soon.

I've been thinking about trying some modifications later on. One thing that might aid turning and movement is adding a fourth servo that turns the robot on the middle point. Looks like I'll be ordering some more servos. 

 

Update 090104:

There, some new videos up. I tried to get more light this time.

I've kept working on the turning and this is as good as I've been able to get it. It's not very efficient but it works. I still think it can be better so I'll keep at it. I've also been able to make it turn while walking forward, but the turn radious is too large for it to be of any use. 

Something needs to be done to the design to make it worth using in the long run. As it is, it can't turn at all on surfaces with high friction. I'd like to try to add a servo in the middle that bends the robot in the horizontal plane to see if it makes it more maneuverable. To try this I'll have to get a new servo and remake the body from scratch. That'll have to wait until I have more time on my hands. in the meantime I'll keep playing around with what I've got.  

 

Comment viewing options

Select your preferred way to display the comments and click "Save settings" to activate your changes.
That's an awesome way to chain up some servos.

Nice - i like the horse hooves clopping

Nice - I don't know if this is an "old & well known" design.. but to me its brand new, and really nice!! For sure this is the quadrapod I'd make if I was to make one now.

I have seen a lot of rather complicated "twists", but I think this one is really good, nice work :) Can you make it turn, I wonder? If so, get some sensors on, and set the beast free :P

PS: Please have more lights on, and film from a lower angle next video.. and get closer to the robot as well, I want to study :)

Cool!

I'll try to get a better video up soon

from the side maybe.

I hope I can make it turn, pretty useless if I can't

great, i love this
So many things on my list. Quadruped is one of them. I'll watch yours develop instead...!
OOooh!! Smooth, tuuurning! :D

 Your robot (and LMR) are feautered on the makezine-blog.

That's cool...

Regarding your thoughts about sensor reading during movement. Even though you robot wobbles and bobs much more than the average "StartHere"-bot, the issue is still the same. You can have the robot stop at regular intervals to do some IR/Ultrasound reading...Maybe do the measuring at the moment when the front and back are flush to get a consistant read?

Designing a rig to mount a cam on, that would get steady pictures during movement; now that would be a cool challenge

 / vzz-clck-"Maneuver"

thanks for pointing that out.

I've thought about placing the sensor on two servos that compensate for the movement. a really cool thing I guess would be to use accelerometers to "sense" a level plane for the sensor. But you're right, the simplest thing is absolutely to let the sensor do its thing while the robot is stationary. I'll order a Sharp IR sensor later this month when I have some more time on my hands. If I order it now I have a feeling that  I won't get anything else done :D

Hey, it appears that Make magazine is kinda following LMR ;D

Make sure to write letsmakerobots.com on (all) your videos if you want people to look by, pay respect and stuff.