The main goal with this project is to get good enough with ROS to continue working on my droid project. The first project is going to be pick and place using an electromagnet: red will go to one side and blue to the other. ROS will be running on a raspberry pi and the steppers will be run off of a cheap CNC control board ontop of an UNO. My plan is to do the motion planning on the pi and send G-Code to the uno. The first hurdle is going to be modeling the motion in ROS, especially because I chose a rather complicated mechanism (where the axes are coupled) so that I could eliminate an extra actuator at the wrist (the end effector stays parallel with the ground at all times.
After that it is going to get much more complicated!
I plan on 3D printing checkers pieces with steel blanks or washers inside of them and having the robot play checkers with me. I designed the arm to be able to reach the length of a standard checkers board, and it should have 6oz of payload capability. I'm planning on using a Pi camera to identify objects and by keeping the geometry of the board consistent, it shouldn't be too insaine to do trajectory calcs and plan the motion. I found some python libraries for a checkers AI, but I am dubious on how predictable I'll be able to get the object detection of the checkers pieces to be. I'm trying to be much more realistic on this project, so I am shooting for feburary 2023 to have something kind of sort of play against me.
Comments