# Solving Inverse Kinematics with unknown orientation

This may be a very basic question, but I haven't had much success finding an answer. I'm currently trying to solve the inverse kinematics of a 3-RRR Planar kinematic arm, whose purpose is to move between points on a graph.

In most theoretical cases the (x,y) coordinates and the orientation of the end effector are known. However, in my case I only know (x,y) coordinates and frankly don't actually care about the orientation of the robot.

My current approach has been to use DH-parameters to solve for each angle. This works, assuming I know the orientation of the end effector, which I do not.

Any help will be greatly appreciated, thank you!

There is no way that you can solve for an IK solution without you -- either explicitly or implicitly -- specifying a criterion for choosing one solution among many others. But that does not mean that solving an IK problem without having any preference on the orientation (or translation) is not possible.

If you solve IK problems analytically, you can just randomly select one orientation and then solve the corresponding IK problem.

If you solve your problems numerically, you already implicitly tell the program which solution to choose by giving the program an initial condition.

• This answer is spot on. Just wanted to give the OP an example in MATLAB of a 3RRR planar manipulator whose IK is solved numerically and whose tip orientation can be taken or not taken into account: it.mathworks.com/matlabcentral/fileexchange/…. Mar 10, 2018 at 9:37