kinematics, improvements

Inverse Kinematics, Briefly:
In setup() I calculate once the position of the output gear on each servo (On). I also calculate the relative position of the joints on the end effector (Pn). When gcode commands arrive I move the end effector matrix, use it to transform the Pn to P’n. Now finding length Do=P’n-On is easy. I use Do, length of rod and length of servo horn to calculate servo angle. Then I tell the servo go to that angle.

I have no working solution for forward kinematics, the math is beyond me. If someone broke it down to a method I could implement then yes, I could do it easily. Anyone want to take on that challenge?

I’ve also been thinking about isolating the code for the rotary component so that a user could #define USE_LINEAR_ACTION or #define USE_ROTARY_ACTION depending on their physical model.

I’d really like to set up a Java app that shows a simulation of the platform in 3D so people can better understand.

Lastly, I’m looking for a killer app – something that showcases what stewart platforms are good for.