0
$\begingroup$

Is it possible to do pick and place tasks using only forward kinematics and object detection information from a camera?

I want to avoid having to do inverse kinematic calculations for my robot. Is there a way I can avoid them for pick and place object sorting?

$\endgroup$
2
  • $\begingroup$ Welcome to Robotics M Fadli Masykuri, but I'm afraid that it is not clear what you are asking. We prefer practical, answerable questions based on actual problems that you face, so it's a good idea to include details of what you want to achieve, what you tried, what you saw & what you expected to see. $\endgroup$
    – Mark Booth
    Nov 22, 2016 at 14:43
  • $\begingroup$ As far as I can see, even if you don't use inverse kinematics to calculate joint positions, you will need to understand the inverse kinematics to move the joints such that you move closer to your intended target, or were you just going just do a random walk? $\endgroup$
    – Mark Booth
    Nov 22, 2016 at 14:44

1 Answer 1

1
$\begingroup$

If you can get the objects to be located consistently at the same positions, you may be able to use "teach and repeat" programming. Otherwise you need some method to translate the objects' positions into joint angles. Inverse kinematics is the most straightforward method. There are some optimization / search algorithms that you could try, too, but they are even more challenging than just computing the closed-form inverse kinematics.

$\endgroup$
4
  • $\begingroup$ thank you Mr, but i still confuse why just Inverse kinematic? what i know is Forward kinematic, the joint angles = the inputs, the outputs = coordinates of the end-effectors. Inverse kinematics, the given inputs = coordinates of the end-effectors, the outputs to calculate are the joint angles,, its right? i use stand kamera behind the robot so i know cordinate the random object and robot, and translate it, i think its work if i just use forward kinematic to pick and place object into box sorting? $\endgroup$ Nov 19, 2016 at 3:26
  • 1
    $\begingroup$ Your phrase "and translate it" means inverse kinematics. $\endgroup$
    – SteveO
    Nov 19, 2016 at 3:27
  • 1
    $\begingroup$ It sounds as if you want to use differential kinematics and iterate until it converges. That is possible to do - however I think closed-form inverse kinematics are much more straightforward to implement. $\endgroup$
    – SteveO
    Nov 19, 2016 at 3:59
  • 1
    $\begingroup$ @M Fadli Masykuri you have pretty much answered it yourself: to do forward kinematics, you need the joint angles. Do you have the joint angles? No. Does your camera give you joint angles? No. $\endgroup$ Nov 19, 2016 at 13:42

Not the answer you're looking for? Browse other questions tagged or ask your own question.