Path Planning and Optimization for Cuspidal 6R Manipulators

2025

Read Path Planning and Optimization for Cuspidal 6R Manipulators (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2501-18505.txt.

Introduction

Alexander Elias and John Wen study how to plan paths for a class of robot arms that can shift between arm poses without passing a bad point. Their paper is from two thousand twenty-five.

Such arms, called cuspidal, are used in trade and have fine form, but planning for them is hard. A path in space may have a valid arm pose at each point, yet no smooth joint path may link them. The team gives a fast way to tell if a robot is of this class, and shows it for some known arms for the first time. They then build a graph of all poses per point and pick the best path by a cost such as joint speed.

Catalog record

YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE56 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2501-18505

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

Path Planning and Optimization for Cuspidal 6R Manipulators Abstract A cuspidal robot can move from one inverse kinematics (IK) solution to another without crossing a singularity. Multiple industrial robots are cuspidal. They tend to have a beautiful mechanical design, but they pose path planning challenges. A task-space path may have a valid IK solution for each point along the path, but a continuous joint-space path may depend on the choice of the IK solution or even be infeasible. This paper presents new analysis, path planning, and optimization methods to enhance the utility of cuspidal robots. We first demonstrate an efficient method to identify cuspidal robots and show, for the first time, that the ABB GoFa and certain robots with three parallel joint axes are cuspidal. We then propose a new path planning method for cuspidal robots by finding all IK solutions for each point along a task-space path and constructing a graph to connect each vertex corresponding to an IK solution. Graph edges have a weight based on the optimization metric, such as minimizing joint velocity. The optimal feasible path is the shortest path in the graph. This method can find non-singular paths as well as smooth paths which pass through singularities. Finally, we incorporate this path planning method into a path optimization algorithm. Given a fixed workspace toolpath, we optimize the offset of the toolpath in the robot base frame while ensuring continuous joint motion. Code examples are available in a publicly accessible repository. \JourName Mechanisms and Robotics \SetAuthorBlock Alexander J. Elias \CorrespondingAuthor email: [email protected] \SetAuthorBlock John T. Wenemail: [email protected] \SetAuthorBlock Department of Electrical, Computer, and Systems Engineering, Rensselaer Polytechnic Institute, Troy, NY 12180, USA keywords Kinematics, Dynamics, and Control of Mechanical Systems; Mechanism Synthesis and Analysis; Mechanisms and Robots; Theoretical and Computational Kinematics 1 Introduction The growing ubiquity of robotic manipulators has resulted in new kinematic designs with different sets of intersecting or parallel joint axes. This has led to the unexpected problem that classical path planning algorithms fail on certain classes of robots due to a property called cuspidality. The most common design for large industrial robot arms features a spherical wrist and two parallel axes. Examples include the ABB IRB 6640 [ 1 ] and the YASKAWA Motoman MA2010-A0 [ 2 ] . Cobots designed for collaborative tasks such as the UR5 [ 3 ] or Techman TM5-700 [ 4 ] feature three parallel axes plus other intersecting axes. In either case, a given end effector pose has up to eight inverse kinematics (IK) solutions which correspond to the solutions of a series of quadratic equations. The IK solutions have intuitive labels and are separated by singularities. Recently, cobots such as the FANUC CRX-10iA/L [ 5 ] (Fig. 1 ) and Kinova Link 6 [ 6 ] (Fig. 1 ) have come to market. These robots have an offset wrist and do not have three parallel axes. This means IK cannot be solved only with quadratics, and there may be more than eight IK solutions. Practitioners have pointed out notable difficulties with path planning and IK for these robots [ 7 ] . The reason for this difficulty is because robots such as the CRX and Link 6 are cuspidal. Cuspidal robots can travel between IK solutions without encountering a singularity. They have IK that cannot be solved with quadratics and have certain constraints on kinematic parameters such as link lengths. For a long time, this property was thought to be impossible. An incorrect proof of this impossibility was published [ 8 ] , but this proof was later proven wrong [ 9 ] . Even today, robot cuspidality is rarely considered [ 7 ] but a robot with a randomly chosen geometry is almost surely cuspidal [ 10 ] . It appears manufacturers and practitioners are building and using cuspidal robots without realizing they are cuspidal or fully appreciating the resulting path planning challenges. Although there is limited research in this field, there has been a recent resurgence in interest because cuspidal manipulators are becoming commercially widespread for the first time. Most research discusses issues arising from manipulators being cuspidal, and many papers conclude cuspidal robots should be avoided. For example, [ 11 ] argued cuspidal robots should be avoided in collaborative applications, and [ 7 ] argued it is reasonable to avoid designing cuspidal robots. {subfigure} [t]0.25 \subcaption {subfigure} [t]0.35 \subcaption {subfigure} [t]0.4 \subcaption Figure 1: Industrial cuspidal robots. (a) FANUC CRX-10iA/L. (b) Kinova Link 6. (c) ABB GoFa CRB 15000 5 kg. The following example illustrates a case where cuspidality can cause path planning issues. Consider the case of robotic welding, which is an advertised use case for many industrial cuspidal robots. A robot operator may program a closed path for a welding operation, and there may be no apparent problem because the robot can perform the entire weld. However, since the robot is cuspidal, the robot can perform a nonsingular change of solution during the welding operation. This means when the robot executes the weld for a second time, the robot may get stuck at a singularity. Moreover, if the first command in the program is a joint-motion command, the robot may produce unexpected large motion to return to the original configuration! For practitioners who are only used to non-cuspidal robots, their existence and many of their properties are nonintuitive. It is difficult to give meaningful names to IK solutions (e.g., elbow up / down) because a robot may move between these solutions with a nonsingular path (meaning they are in the same aspect ). Furthermore, there may be smooth task-space paths such as straight-line Cartesian MoveL commands with IK solutions entirely in one aspect that cannot be smoothly followed. This is because local boundary singularities prevent end effector movement within the workspace. In general, these singularities form surfaces that cannot be avoided with small path deviations, unlike, say, a wrist singularity in a conventional industrial robot with a spherical wrist. In some cases, a closed Cartesian path can be followed once, but attempting to follow the path a second time results in the robot getting stuck at a singularity because following the path causes the robot to switch IK solutions. Cuspidal robots highlight the importance of considering all IK solutions during path planning and demonstrate the need for IK solvers which can find all IK solutions. Missing IK solutions is problematic because feasible paths may be missed, resulting in a failure to connect two points or picking a suboptimal path. (In fact, finding all IK solutions is important for path planning even for noncuspidal robots as certain paths may be non-optimal or even infeasible depending on the choice of initial IK solution.) Many IK solvers are Jacobian-based iterative solvers, and these do not return all IK solutions. IK solvers used by industrial robots which return multiple solutions may have problems. ROBOGUIDE, the software used by FANUC to control their CRX line of robots, only provides up to eight IK solutions even though the CRX has up to sixteen in general [ 10 ] . Furthermore, ROBOGUIDE labels IK solutions using shoulder/elbow/wrist configuration parameters, but these descriptions may not correctly match the configuration of the arm and may not change when changing

Read the complete text