felixmaisonneuve
commented
3 years ago Hi @atlas-jj,
Unfortunately, we do not provide our Jacobian matrix.
You can use the third party library Robotics Library. However, we use a custom calibration to get a highier degree of precision that you won't be able to achieve using this library.
I can also link this article that explains the algorithm used by our API : Phuoc, L. M., Lee, S. H., Kim, H. M., & Martinet, P. (2008). Damped least square based genetic algorithm with Gaussian distribution of damping factor for singularity-robust inverse kinematics. Journal of Mechanical Science and Technology, 22(7), 1330-1338.
I will close this issue, but if you have any more questions feel free to post them and I will answer them
Regards, Felix
Summary
Is there any easy way to get the GEN3 Jacobian matrix just like the other robots like Franka Panda and WAM?
Use case
Why is this feature needed?
Well, Jacobian matrix defines dot(e) = J * dot(q), it comes inherently from the robot's geometry model, it will be useful in many tasks like visual serving or reinforcement learning if RL practitioners care about more efficient training scenarios.
Alternative
Should be best if provided within your kortex Python API.
If this feature is not implemented, do you have a workaround?
Nope.
Possible issues
N/A
Additional context
I'm running my thesis experiment on GEN3, so that's quite urgent. Will be appreciated if you can guide me to get sufficient parameters to compute the Jacobian.