Legacy Forum: Preserving Nearly 20 Years of Community History - A Time Capsule of Discussions, Memories, and Shared Experiences.

Robonova walking

Hitec robotics including ROBONOVA humanoid, HSR-8498HB servos, MR C-3024 Controllers and RoboBasic
4 postsPage 1 of 1
4 postsPage 1 of 1

Robonova walking

Post by beermat » Wed Aug 02, 2006 3:58 am

Post by beermat
Wed Aug 02, 2006 3:58 am

When looking at videos at KHR-* walking, both bots seem much more adept and 'confident' at walking forward quickly than the Robonova. The stock walking moves for RN always seem very timid to me (I dont count the 'fast walk' RN stock move, that's not that great or stable for propelling RN forward).

Is this a design issue / COG issue for the Robonova, or a code issue, with just very conservative code out of the box? Both KHRs dont appear to have or use any additional DOF for their walking motion. Also, I dont recall seeing anyone in the community improving on the stock walk; maybe because handsprings are so much cooler to work on? :)

Anyone have any thoughts on this?
When looking at videos at KHR-* walking, both bots seem much more adept and 'confident' at walking forward quickly than the Robonova. The stock walking moves for RN always seem very timid to me (I dont count the 'fast walk' RN stock move, that's not that great or stable for propelling RN forward).

Is this a design issue / COG issue for the Robonova, or a code issue, with just very conservative code out of the box? Both KHRs dont appear to have or use any additional DOF for their walking motion. Also, I dont recall seeing anyone in the community improving on the stock walk; maybe because handsprings are so much cooler to work on? :)

Anyone have any thoughts on this?
beermat
Savvy Roboteer
Savvy Roboteer
User avatar
Posts: 137
Joined: Sun Jul 23, 2006 1:00 am

Post by Ray » Wed Aug 02, 2006 5:09 am

Post by Ray
Wed Aug 02, 2006 5:09 am

Ya, I believe the mechanical design of KHR is better,
but the "confident" motion might have included the installation of gyros.

so, you should install the gyros also, before comparing again :D
Ya, I believe the mechanical design of KHR is better,
but the "confident" motion might have included the installation of gyros.

so, you should install the gyros also, before comparing again :D
Ray
Savvy Roboteer
Savvy Roboteer
User avatar
Posts: 230
Joined: Sun Apr 23, 2006 1:00 am
Location: HK

Post by Robo1 » Wed Aug 02, 2006 12:09 pm

Post by Robo1
Wed Aug 02, 2006 12:09 pm

I've got a khr-2 and I can say the standard walk isn't like the demo. I've had it for 2 weeks and can get it to walk well but have to slow down all the speed setting otherwise it looks like a made man on crack and vibrates and shakes every where. I think the ones in hte demo have gyros which would help.

bren
I've got a khr-2 and I can say the standard walk isn't like the demo. I've had it for 2 weeks and can get it to walk well but have to slow down all the speed setting otherwise it looks like a made man on crack and vibrates and shakes every where. I think the ones in hte demo have gyros which would help.

bren
Robo1
Savvy Roboteer
Savvy Roboteer
Posts: 501
Joined: Fri Jun 30, 2006 1:00 am
Location: UK - Bristol

Post by hivemind » Wed Aug 02, 2006 3:04 pm

Post by hivemind
Wed Aug 02, 2006 3:04 pm

ive worked on walking here and there, but not nearly as much as i could... I am waiting for my new pieces to be biult beforei do so because that might be my final design, with the independant leg rotation and all.

If you want my advice, it would be ti put the servos on highspeed, tilt both feet outwards slightly (to a very balanced position, but not too much so) and then pop them both up and down, but not very much. You should work in places until this code works, and then have him lean forward ever so slightly, and it will turn into a stable run, similar to the japanese home-built robots. It is all about the little completely stable steps.

I have tried large steps, where i had my robot going through some karate stances (I happen to be a black belt in shotokan, I've been doing it for a long time, so I have gotten a sense of how I ease into these stances) and essentially the walk would have some weaker points and some stronger points, and i decided that was no good and that watching the robots run on all the videos i had seen, it would be best to have smaller steps...

Anyway, thats my advice for anyone who would like to try a walking code, its a fun thing to try, being so essential and all. goodluck! :)
ive worked on walking here and there, but not nearly as much as i could... I am waiting for my new pieces to be biult beforei do so because that might be my final design, with the independant leg rotation and all.

If you want my advice, it would be ti put the servos on highspeed, tilt both feet outwards slightly (to a very balanced position, but not too much so) and then pop them both up and down, but not very much. You should work in places until this code works, and then have him lean forward ever so slightly, and it will turn into a stable run, similar to the japanese home-built robots. It is all about the little completely stable steps.

I have tried large steps, where i had my robot going through some karate stances (I happen to be a black belt in shotokan, I've been doing it for a long time, so I have gotten a sense of how I ease into these stances) and essentially the walk would have some weaker points and some stronger points, and i decided that was no good and that watching the robots run on all the videos i had seen, it would be best to have smaller steps...

Anyway, thats my advice for anyone who would like to try a walking code, its a fun thing to try, being so essential and all. goodluck! :)
Image
hivemind
Savvy Roboteer
Savvy Roboteer
User avatar
Posts: 211
Joined: Sat Jul 01, 2006 1:00 am
Location: between my computer and robot.


4 postsPage 1 of 1
4 postsPage 1 of 1