SANDROS: a motion planner with performance proportional to task difficulty

Abstract

To address the need of a practical motion planner for manipulators, we present an efficient and resolution-complete algorithm that has performance commensurate with task difficu._ty. The algorithm uses SANDROS, a new search strategy that combines hierarchical, nonuniform-multi-resolution, and be_t-first search to find a near-optimal solution in the configuration space. This algorithm can be applied to any manipulator, and has been tested with 5 and 6-degree-of-freedom robots, with execution time ranging from 20 seconds to 10 minutes on a 16 MIPS workstation.

Combinatorial Motion Planning

“…Either a standard grid or a Sukharev grid can be used. Let 35) in which the first i − 1 components and the last n − i components are 0. A grid point is a configuration q ∈ C that can be expressed in the form 10 36) in which each j i ∈ {0, 1, .…”

“…Some aspects of sampling-based motion planning are still evolving. Early approaches include [14,25,35,48,49,52,53,125,138]. The Gilbert-Johnson-Keerthi algorithm [62] is an early collision detection approach that helped inspire sampling-based motion planning; see [80] and [102] for many early references.…”

Learning to improve path planning performance

“…As mentioned in the introduction, a large amount of research has been done on robot path planning, most of which deals with solving one-time problems in stationary environments [2, 3, 6,15,20,22,271. Most implemented path planners have been developed for mobile robots and manipulators with a few degrees of freedom (dof).…”

Planning Movement of a Robotic Arm for Assembly of Products

“…However, the computational time is still too high to be appreciated in real-time applications. For example, in the case of 4 D DAVID PUBLISHING degrees of freedom manipulator, if each degree of freedom is quantified in 90 discrete levels, then the C-space of the robotic arm will be 904 = 6.5 × 107 configurations and the manipulator of six unions will need more than 500 billions of configurations [9,10], it indicates that for a six degrees of freedom manipulator, the C-space bitmap with a resolution of 9° and 360° in each union, would need about 512 MB of memory.…”