Learning in robotic task planning

Planning has been an intricate part of robotic development beginning with the fixed programs of the zero generation manipulator robots and the first generation programmable robots. The second generation adaptive robots began the conventional planning by incorporating sensor input and appropriate planning technics. However, with the advancements in sensors and microcomputer capabilities, the third generation robot ushered in an era of improved planning technics for robotics (Vukobratovic, et al., 1988).