Rapidly-Exploring Quotient-Space Trees: Motion Planning Using Sequential Simplifications

Motion planning problems can be simplified by admissible projections of the configuration space to sequences lower-dimensional quotient-spaces, called sequential simplifications. To exploit simplifications, we present Quotient-space Rapidly-exploring Random Trees (QRRT) algorithm. QRRT takes as input a start and goal configuration, sequence quotient-spaces. The algorithm grows trees on quotient-spaces both sequentially simultaneously guarantee dense coverage. is shown (1) probabilistically complete, (2) reduce runtime at least one order magnitude. However, show in experiments that varies substantially between different quotient-space sequences. find out why, perform an additional experiment, showing more narrow environment, runtime.