Darwin-OP code
Darwin-OP code
by icecream
Sat Nov 17, 2012 2:25 pm
Hi, what is Darwin-OP code? What so special about it and the Darwin framework when compared with usb2dynamixel and RobotPlus?
by PedroR
Mon Nov 19, 2012 11:26 am
Hi icecream
DarWin-Op has two components: the hardware design and the Software Framework
The software library can be ported to any other humanoid and/or robot design with minimal effort provided you maintain the same mechanical structure (ie the positiong of the servos).
The dimensions can change and you can easily adjust this in code.
The Framework itself - which is what I believe is your question - has great features for making coding for DARwIn-Op easier:
1) It includes an automatic gait generator.
This means you only need to tell the librayr to "walkTo", give the necessary parameters and ti will automatically make the RObot walk and adjust the walk dynamically based on the input from the IMU
2) It includes a number of additional features including a Vision API to make BLOB tracking very easy.
3) It can also play static/pre recorded Roboplus Motion files
4) Because the Framework assumes you have a PC onboard the Robot it expects to talk directly to the servos "as if" you had a USBDynamixel.
(you should actually be able to run the framework from your PC if you have a USb2Dynamixel although the gait generator needs the IMU readings)
In reality, in DARwIn-OP robot there is an additional device - the CM730 - which acts as both a USb2Dynamixle and as a Dynamixel device itself (with its own ID).
As a Dynamixel device the CM-730 provides IMU readings and some functionality related to HRI (reading button, controlling eye and forehead LEDs, etc).
So in short, the main difference between the Framework and the plain USb2Dynamixel/Roboplus Motion is that the Framework is capable of automatic walk generation and additional libraries for Vision, etc.)
Pedro.
DarWin-Op has two components: the hardware design and the Software Framework
The software library can be ported to any other humanoid and/or robot design with minimal effort provided you maintain the same mechanical structure (ie the positiong of the servos).
The dimensions can change and you can easily adjust this in code.
The Framework itself - which is what I believe is your question - has great features for making coding for DARwIn-Op easier:
1) It includes an automatic gait generator.
This means you only need to tell the librayr to "walkTo", give the necessary parameters and ti will automatically make the RObot walk and adjust the walk dynamically based on the input from the IMU
2) It includes a number of additional features including a Vision API to make BLOB tracking very easy.
3) It can also play static/pre recorded Roboplus Motion files
4) Because the Framework assumes you have a PC onboard the Robot it expects to talk directly to the servos "as if" you had a USBDynamixel.
(you should actually be able to run the framework from your PC if you have a USb2Dynamixel although the gait generator needs the IMU readings)
In reality, in DARwIn-OP robot there is an additional device - the CM730 - which acts as both a USb2Dynamixle and as a Dynamixel device itself (with its own ID).
As a Dynamixel device the CM-730 provides IMU readings and some functionality related to HRI (reading button, controlling eye and forehead LEDs, etc).
So in short, the main difference between the Framework and the plain USb2Dynamixel/Roboplus Motion is that the Framework is capable of automatic walk generation and additional libraries for Vision, etc.)
Pedro.
