Minimally actuated hyper-redundant robots: Motion planning methods based on fractals and self-organizing systems

Moshe P. Mann, Lior Damti, David Zarrouk

Research output: Contribution to journalArticlepeer-review

2 Scopus citations

Abstract

This article presents a motion planning method for a novel hyper-redundant robot with minimal actuation based on the principles of fractals and self-organizing systems. The robot consists of multiple links connected by passive joints and a movable actuator. The actuator travels over the links to a given joint and adjusts the relative angle between the two adjacent links allowing the robot to undergo the same wide range of motions of hyper-redundant robot but with only one actuator. A suitable objective of the motion planner is to minimize the number of actuator traversals, which translates into minimizing the number of bends in the c-space trajectory. To this end, we propose a novel method for motion planning using fractals and self-organizing systems. A self-similar pattern for the path is implemented to map a path from start to finish. Each iteration of path segments is of smaller dimension than the previous one and is appended to it, just as in classical fractals. This process continues until a feasible trajectory is calculated. Self-organizing systems are then applied to this trajectory post-processing to optimize it by eliminating bends in the path. Examples of the robot maneuvering around obstacles and through confined spaces are shown to demonstrate the efficacy of the motion planner.

Original languageEnglish
JournalInternational Journal of Advanced Robotic Systems
Volume16
Issue number2
DOIs
StatePublished - 1 Mar 2019

Keywords

  • Hyper-redundant robot
  • L-systems
  • fractal
  • minimally actuated
  • non-holonomic motion planning
  • self-organizing system

ASJC Scopus subject areas

  • Software
  • Computer Science Applications
  • Artificial Intelligence

Fingerprint

Dive into the research topics of 'Minimally actuated hyper-redundant robots: Motion planning methods based on fractals and self-organizing systems'. Together they form a unique fingerprint.

Cite this