Automatic generation of goal regions for assembly tasks in the presence of uncertainty

A systematic procedure for generating the goal region of a mating action from a CAD model of assemblies using the mating features and the constraint types residing in the model, is presented. Assuming convexity and similarity of mating features, an analytical solution for C-space interior is developed and used to construct the goal region from 2-D features. This method provides a well-defined goal region which allows the computation of the corresponding success probability. Due to the uncertainties in object representation, goal regions are also subject to uncertainties. An approach that shrinks the nominal boundary of a goal region is proposed for computing the expectation of success probabilities. The effects of interference and fine-motion planning on success probabilities are also taken into consideration in order to define theoretical goal regions