# Lecture 7 ## Movie segment: volley boll playin robot. * what is Zi ?how to calculate it? : it is basically z axis or direction of angular velocity vector. * ![](https://i.imgur.com/ECs4ZxG.png) * Here we are finding end effector linear and angular velocity. * angular velocity for only prismatic joints is zero for end-effector. * ![](https://i.imgur.com/3TIuESy.png) * here we have substituted Ziqdot to get below expressions.. * What is E1 , E2 .. and E1dash? * This Jw and Jv are made up of Z1,Z2... * Here we haven't decided any frame in which we have decided vector. * Jv and Jw can be also simplified further if we have computed forward kinematics. * if we use cartesian co-ordinates (instead of spherical co-ordiantes) we can directly differentiate Xp to get Jv. ![](https://i.imgur.com/EJRtAGi.png) * How we will do differentiation of Xp w.r.t q? * In second slide below we have Zi in frame i then we convert it to frame zero(0) with rotation matrix. * So basically we find oZ1, 0Z2 in last column of rotation matrix 0R1,0R2 etc. ![](https://i.imgur.com/wdW7LAe.png) * here he has determined J from basics i.e, from E1,E2 wala formula. ![](https://i.imgur.com/QAW2orq.png) * ![](https://i.imgur.com/s1YofBB.png) * Now we are simply finding translational matrix ![](https://i.imgur.com/JqJlUVC.png) * we need all the vectors in frame zero. * Here we need only 3rd column of matrix. * the red rectangles are Z1, Z2 etc.in frame zero. * He mentioned that he will not find 2T3 then multiply with 1T2 to get 1T3 he will directly find 1T3 but he didn't explained how?..[00:51:58](https://youtu.be/6SRTAoyzC6A?t=3118) : well he was mentioning about order of multiplication i.e, go from 0 to 6 and not reverse. * the last colomn is same from matrix 3 because we have reached end-effector. * We simply partially derivate the last column (as it represent the position of end-effector) to get ${J}_v$ * ![](https://i.imgur.com/f7DYgWe.png) * Below we have got jacobian for scheinman arm ![](https://i.imgur.com/y9E4Qf2.png) * At singularity two columns in jacobian are going to be dependent and also matrix is going to lose rank and we like to be full matrix. # Kinematic singularity: ![](https://i.imgur.com/X7uxKYf.png)