ABSTRACT
It is well known that manufacturing cost of robots increase rapidly as the tolerances on the components are made tighter. The allocation of proper tolerances is therefore one of the most important tasks if the finished design is to achieve its intended purpose and yet be economical to produce. In order to minimize the manufacturing process cost, we propose a least-cost tolerance method used at the design stage of robots. In particular, pseudo-boolean programming is used as a method to allocate tolerances to robot parameters. Statistical approach rather than arithmetic approach is adopted in tolerance and error analysis. By Denavit-Hartenberg representation and the first order approximation of differential changes, the position and orientation errors of the end effector are estimated statistically. Techniques to reduce the problem size are also provided. In the proposed model, position and orientation errors of the end effector become constraints to be met. The model is applied to a revolute joint with six degrees of freedom.