Motion Design through Inverse Kinematics

From the literature, I understand that inverse kinematics can be obtained analytically, if the chain has five or less DOF. If the chain has six DOF, an analytical solution can be obtained, only if three consecutive joints have intersecting axes (e.g NAO hip). There are 4 DOF at the two shoulder-elbow chain , 3 DOF at head-torso chain. The legs have six DOF, however the meet the condition given above because the three hip joints have intersecting axes. Therefore, it is possible to obtain a fully analytical solution for the inverse kinematics of the NAO.


Aldebaran provides the DH parameters for all the joints of the NAO robot and that was very useful. It also provides an inverse kinematics mechanism integrated within the proprietary NaoQi middleware for the NAO robot. These functions in the API of the robot can move the end effector of a kinematic chain to a given point in the three-dimensional space. Since it is using the iterative approach, singularities might occur more often than desired. However, as the goal is to design finite number of reflex movements, the Aldebaran API can be considered.

To be continued...