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

darwin motions

Dedicated to the new Dynamic Anthropomorphic Robot with Intelligence - Open Platform By Robotis and related projects
45 postsPage 1 of 31, 2, 3
45 postsPage 1 of 31, 2, 3

darwin motions

Post by robaldo » Sun Apr 10, 2011 9:35 am

Post by robaldo
Sun Apr 10, 2011 9:35 am

Hi all,
where can i find DARwin's motion files?
did anybody work with DARwin?
could you please tell me how DARwin is doing motions?
thanks
Hi all,
where can i find DARwin's motion files?
did anybody work with DARwin?
could you please tell me how DARwin is doing motions?
thanks
robaldo
Savvy Roboteer
Savvy Roboteer
Posts: 53
Joined: Fri Dec 03, 2010 4:49 pm

Post by i-Bot » Sun Apr 10, 2011 11:30 am

Post by i-Bot
Sun Apr 10, 2011 11:30 am

If you look at the posts here, you will see bullit has a DARwin.

http://robosavvy.com/forum/viewtopic.ph ... sc&start=0

If you set up the experiment I describe, you will see the DARwin motions in Roboplus. The source code is all there to see how they are interpreted.

DARwin can do motions under Roboplus, but also has additional new walking code based on IK with a gait state machine and gyro stabilisation. Walk parameters can be tuned.
If you look at the posts here, you will see bullit has a DARwin.

http://robosavvy.com/forum/viewtopic.ph ... sc&start=0

If you set up the experiment I describe, you will see the DARwin motions in Roboplus. The source code is all there to see how they are interpreted.

DARwin can do motions under Roboplus, but also has additional new walking code based on IK with a gait state machine and gyro stabilisation. Walk parameters can be tuned.
i-Bot
Savvy Roboteer
Savvy Roboteer
User avatar
Posts: 1142
Joined: Wed May 17, 2006 1:00 am

Post by robaldo » Sun Apr 10, 2011 2:21 pm

Post by robaldo
Sun Apr 10, 2011 2:21 pm

Hi i-Bot,
thanks for reply
i have bioloid premium and ax-12s
i have controlled them before with usb2dynamixel and my laptop without cm510, now i want to use darwin code with them...
i want you to tell me, how can i use darwin code and where is the code to use it to control ax-12s with darwin code?
i almost misunderstood of what you shared before...
sorry
Hi i-Bot,
thanks for reply
i have bioloid premium and ax-12s
i have controlled them before with usb2dynamixel and my laptop without cm510, now i want to use darwin code with them...
i want you to tell me, how can i use darwin code and where is the code to use it to control ax-12s with darwin code?
i almost misunderstood of what you shared before...
sorry
robaldo
Savvy Roboteer
Savvy Roboteer
Posts: 53
Joined: Fri Dec 03, 2010 4:49 pm

Post by i-Bot » Sun Apr 10, 2011 5:40 pm

Post by i-Bot
Sun Apr 10, 2011 5:40 pm

In the Robotis DARwin code the on board linux controller and subcontroller combo replace the CM510. There is no Windows version yet, but this is something which could be created, so the Darwin controller program runs as a task under Windows. Roboplus could talk to this via TCP/IP socket on same PC.
In the Robotis DARwin code the on board linux controller and subcontroller combo replace the CM510. There is no Windows version yet, but this is something which could be created, so the Darwin controller program runs as a task under Windows. Roboplus could talk to this via TCP/IP socket on same PC.
i-Bot
Savvy Roboteer
Savvy Roboteer
User avatar
Posts: 1142
Joined: Wed May 17, 2006 1:00 am

Post by robaldo » Sun Apr 10, 2011 5:58 pm

Post by robaldo
Sun Apr 10, 2011 5:58 pm

i-Bot wrote:In the Robotis DARwin code the on board linux controller and subcontroller combo replace the CM510. There is no Windows version yet, but this is something which could be created, so the Darwin controller program runs as a task under Windows. Roboplus could talk to this via TCP/IP socket on same PC.

now i don't want to use the roboplus motion or something like that
i have downloaded the darwin code and find some files like walking.h walking.cpp and so on
i want to know whether i can use them wih ax12 or not?
i am using both windows and linux.
i-Bot wrote:In the Robotis DARwin code the on board linux controller and subcontroller combo replace the CM510. There is no Windows version yet, but this is something which could be created, so the Darwin controller program runs as a task under Windows. Roboplus could talk to this via TCP/IP socket on same PC.

now i don't want to use the roboplus motion or something like that
i have downloaded the darwin code and find some files like walking.h walking.cpp and so on
i want to know whether i can use them wih ax12 or not?
i am using both windows and linux.
robaldo
Savvy Roboteer
Savvy Roboteer
Posts: 53
Joined: Fri Dec 03, 2010 4:49 pm

Post by i-Bot » Sun Apr 10, 2011 8:29 pm

Post by i-Bot
Sun Apr 10, 2011 8:29 pm

I'm using them with AX12. Though RX-28M will have greater capability eventually, this is not exploited yet in published code.
I'm using them with AX12. Though RX-28M will have greater capability eventually, this is not exploited yet in published code.
i-Bot
Savvy Roboteer
Savvy Roboteer
User avatar
Posts: 1142
Joined: Wed May 17, 2006 1:00 am

Post by Fritzoid » Mon Apr 11, 2011 11:01 am

Post by Fritzoid
Mon Apr 11, 2011 11:01 am

You will need to update darwinop/trunk/darwin/Framework/include/Kinematics.h to use walk.cpp with a Bioloid.
You will need to update darwinop/trunk/darwin/Framework/include/Kinematics.h to use walk.cpp with a Bioloid.
Fritzoid
Savvy Roboteer
Savvy Roboteer
User avatar
Posts: 331
Joined: Mon Dec 18, 2006 1:00 am

Post by i-Bot » Mon Apr 11, 2011 11:57 am

Post by i-Bot
Mon Apr 11, 2011 11:57 am

This diagram helps to understand the values in kinematics.h

http://sourceforge.net/projects/darwinop/files/Hardware/Mechanics/Physical%20Information/DARwIn-OP_Kinematics.pdf/download

Note also that the elbow is at right angles on DARwin, swinging front to back The walking arm swing is from the shoulder, not the elbow, but you may need to adjust some initial value settings in walking.cpp.
This diagram helps to understand the values in kinematics.h

http://sourceforge.net/projects/darwinop/files/Hardware/Mechanics/Physical%20Information/DARwIn-OP_Kinematics.pdf/download

Note also that the elbow is at right angles on DARwin, swinging front to back The walking arm swing is from the shoulder, not the elbow, but you may need to adjust some initial value settings in walking.cpp.
i-Bot
Savvy Roboteer
Savvy Roboteer
User avatar
Posts: 1142
Joined: Wed May 17, 2006 1:00 am

Post by Fritzoid » Wed Apr 13, 2011 12:01 pm

Post by Fritzoid
Wed Apr 13, 2011 12:01 pm

I just ran a simulation using the DarwinOP walking_tuner motions with a Bioloid premium model. The results were not good! It turns out that the Darwin kinematics are different than the type A in significant ways.

The most obvious problem is the hips. The hip brackets (F5) on the Bioloid are mounted at a 45-degree angle, whereas the DarwinOP has them centered (more like a comprehensive).

The lower arms are mounted at 90-degrees relative to the upper arms on the Darwin. It's an improvement over the Bioloid design, but different none the less.

What I call the groin (servos 9 and 10) are mounted facing forward on Darwin and backward on the Bioloid. This is probably the case with the foot servos (17 and 18) too.

There also seems to be a problem with the knee swing.

In conclusion, I recommend you think twice before trying this with a real robot!
I just ran a simulation using the DarwinOP walking_tuner motions with a Bioloid premium model. The results were not good! It turns out that the Darwin kinematics are different than the type A in significant ways.

The most obvious problem is the hips. The hip brackets (F5) on the Bioloid are mounted at a 45-degree angle, whereas the DarwinOP has them centered (more like a comprehensive).

The lower arms are mounted at 90-degrees relative to the upper arms on the Darwin. It's an improvement over the Bioloid design, but different none the less.

What I call the groin (servos 9 and 10) are mounted facing forward on Darwin and backward on the Bioloid. This is probably the case with the foot servos (17 and 18) too.

There also seems to be a problem with the knee swing.

In conclusion, I recommend you think twice before trying this with a real robot!
Fritzoid
Savvy Roboteer
Savvy Roboteer
User avatar
Posts: 331
Joined: Mon Dec 18, 2006 1:00 am

Post by PedroR » Wed Apr 13, 2011 5:47 pm

Post by PedroR
Wed Apr 13, 2011 5:47 pm

Hi Fritzoid, i Bot

We at RoboSavvy are all really excited about the work you've been doing on this port of DarwinOp to Bioloid (and I'm sure plenty of other members as well).

I was wondering if you could post some videos of your attempts or work so far.

Even if they're not functional it'd be really interesting to see how far along you are at this point :)

Thanks and I look forward for more news!
Pedro.
Hi Fritzoid, i Bot

We at RoboSavvy are all really excited about the work you've been doing on this port of DarwinOp to Bioloid (and I'm sure plenty of other members as well).

I was wondering if you could post some videos of your attempts or work so far.

Even if they're not functional it'd be really interesting to see how far along you are at this point :)

Thanks and I look forward for more news!
Pedro.
PedroR
Savvy Roboteer
Savvy Roboteer
Posts: 1199
Joined: Mon Jun 16, 2008 11:07 pm

Post by Fritzoid » Mon May 02, 2011 12:30 pm

Post by Fritzoid
Mon May 02, 2011 12:30 pm

I made the necessary modifications to the DarwinOP walking code so that it can be used with a Bioloid Premium. Beyond the leg link sizes and some reversed servos (mentioned above) it seems that there is something strange about the shoulders on the Darwin. I had to make a 45-degree correction for them as well as the hip servos. Here's a video of a reasonably stable walk.

phpBB [media]
I made the necessary modifications to the DarwinOP walking code so that it can be used with a Bioloid Premium. Beyond the leg link sizes and some reversed servos (mentioned above) it seems that there is something strange about the shoulders on the Darwin. I had to make a 45-degree correction for them as well as the hip servos. Here's a video of a reasonably stable walk.

phpBB [media]
Fritzoid
Savvy Roboteer
Savvy Roboteer
User avatar
Posts: 331
Joined: Mon Dec 18, 2006 1:00 am

Post by limor » Mon May 02, 2011 5:35 pm

Post by limor
Mon May 02, 2011 5:35 pm

Hi Fritzoid
Can you please post the bioloid premium rigged model and modified Darwin code ?

I would like to try and port this to an opensource robot simulator that we have been developing ezphysics.org. This simulator is designed to develop closed loop control systems for robots. See http://actuated.wordpress.com for more details.


thanks
Limor
Hi Fritzoid
Can you please post the bioloid premium rigged model and modified Darwin code ?

I would like to try and port this to an opensource robot simulator that we have been developing ezphysics.org. This simulator is designed to develop closed loop control systems for robots. See http://actuated.wordpress.com for more details.


thanks
Limor
limor
Savvy Roboteer
Savvy Roboteer
User avatar
Posts: 1845
Joined: Mon Oct 11, 2004 1:00 am
Location: London, UK

Post by Fritzoid » Tue May 03, 2011 12:23 pm

Post by Fritzoid
Tue May 03, 2011 12:23 pm

Hi limor,

I'm not sure how applicable the simulator stuff will be to your project but I am happy to pass along what I have. And all I have so far is some disjoint pieces. There isn't any documentation either because I'm really still in the design phase. That said, I've stored an archive in my builders area that you might find interesting.

http://robosavvy.com/Builders/Fritzoid/BioSim.zip

I have followed your actuated character project with interest for some time now. It's a much more ambitious project than my own and you have obviously made good progress so far. Keep up the good work.
Hi limor,

I'm not sure how applicable the simulator stuff will be to your project but I am happy to pass along what I have. And all I have so far is some disjoint pieces. There isn't any documentation either because I'm really still in the design phase. That said, I've stored an archive in my builders area that you might find interesting.

http://robosavvy.com/Builders/Fritzoid/BioSim.zip

I have followed your actuated character project with interest for some time now. It's a much more ambitious project than my own and you have obviously made good progress so far. Keep up the good work.
Fritzoid
Savvy Roboteer
Savvy Roboteer
User avatar
Posts: 331
Joined: Mon Dec 18, 2006 1:00 am

Post by limor » Tue May 03, 2011 1:20 pm

Post by limor
Tue May 03, 2011 1:20 pm

Thanks for sharing!

Are all the parts of the robot in separate wrl files?
Does the humanoid_a contain the skeleton and all the meshes in one place?
Where are the physics parameters (weight, intertial tensor) ?
Thanks for sharing!

Are all the parts of the robot in separate wrl files?
Does the humanoid_a contain the skeleton and all the meshes in one place?
Where are the physics parameters (weight, intertial tensor) ?
limor
Savvy Roboteer
Savvy Roboteer
User avatar
Posts: 1845
Joined: Mon Oct 11, 2004 1:00 am
Location: London, UK

Post by Fritzoid » Tue May 03, 2011 3:27 pm

Post by Fritzoid
Tue May 03, 2011 3:27 pm

The physical parameters are defined in the file humanoid_A.wrl. That main file references the various other VRML mesh files for 3D display and collision detection purposes. The joint hierarchy along with the link definitions can all be found in the main file which can be easily viewed.

The COG and inertial tensor parameters were extracted from the set of high-resolution parts and assemblies that I created from the Robotis igs files. These parts and assemblies are in Alibre Design format and I can make them available too if you are interested.
The physical parameters are defined in the file humanoid_A.wrl. That main file references the various other VRML mesh files for 3D display and collision detection purposes. The joint hierarchy along with the link definitions can all be found in the main file which can be easily viewed.

The COG and inertial tensor parameters were extracted from the set of high-resolution parts and assemblies that I created from the Robotis igs files. These parts and assemblies are in Alibre Design format and I can make them available too if you are interested.
Fritzoid
Savvy Roboteer
Savvy Roboteer
User avatar
Posts: 331
Joined: Mon Dec 18, 2006 1:00 am

Next
45 postsPage 1 of 31, 2, 3
45 postsPage 1 of 31, 2, 3