0
$\begingroup$

Rosanswers logo

I would like to use a robotarm to pick up objects that are moved by a conveyor belt.

The way I would go about this if the object was stationary is using vision to detect the object, then create a 3D file of that object and use that to place the object in the environment of the robotic arm, then with MoveIt! functionality pick up the object.

Is there a (simple) way to expand upon this method to pick up objects from a moving conveyor belt with encoder?

I could calculate where the object would be after a few seconds and time the movements of the robot to line up with that. But that would be very sensitive to errors or changes in the speed of the belt. I would rather make the robotarm somewhat 'smarter' and able to adjust to the movements of the conveyor belt in a somewhat real-time manner.

Any ideas?

Thank you for your help!

Originally posted by Gengoz on ROS Answers with karma: 45 on 2015-07-14

Post score: 1

Improve this question
$\endgroup$

1 Answer 1

Reset to default
0
$\begingroup$

Rosanswers logo

I guess you should be able to set the goal pose of the gripper in a tf-frame that you attach to the belt. You could then read the encoder of the conveyor and update the transformation to the frame accordingly. Moveit should then replan the grasp according to the slightly moved goal-tf so that you should be able to change the speed of the conveyor after your first detection and still be able to grasp the object.

Originally posted by NEngelhard with karma: 3519 on 2015-07-14

This answer was ACCEPTED on the original site

Post score: 1

Improve this answer
$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.