Rapidly Exploring Random Tree Algorithm-Based Path Planning for Worm-Like Robot.

Yifan Wang,Zehao Liu,Kathryn A Daltorio,Akhil Kandhari,Prathamesh Pandit

doi:10.3390/biomimetics5020026

Yifan Wang, Zehao Liu + Show 3 more

Open Access

https://doi.org/10.3390/biomimetics5020026

Copy DOI

Abstract

Inspired by earthworms, worm-like robots use peristaltic waves to locomote. While there has been research on generating and optimizing the peristalsis wave, path planning for such worm-like robots has not been well explored. In this paper, we evaluate rapidly exploring random tree (RRT) algorithms for path planning in worm-like robots. The kinematics of peristaltic locomotion constrain the potential for turning in a non-holonomic way if slip is avoided. Here we show that adding an elliptical path generating algorithm, especially a two-step enhanced algorithm that searches path both forward and backward simultaneously, can make planning such waves feasible and efficient by reducing required iterations by up around 2 orders of magnitude. With this path planner, it is possible to calculate the number of waves to get to arbitrary combinations of position and orientation in a space. This reveals boundaries in configuration space that can be used to determine whether to continue forward or back-up before maneuvering, as in the worm-like equivalent of parallel parking. The high number of waves required to shift the body laterally by even a single body width suggests that strategies for lateral motion, planning around obstacles and responsive behaviors will be important for future worm-like robots.

Highlights

Due to soft characteristics, nonholonomic constraints, limits on reachable space and the high number of degrees of freedom (DOF), navigating and path planning for worm-like robots can be difficult [1,2,3]
Worm-like robots locomote by changing the body shape of each segment
We introduced a more advanced algorithm (enhanced combined rapidly exploring random tree (RRT) ellipse (Algorithm 4)) with some helpful improvements

Summary

Introduction

Nonholonomic constraints, limits on reachable space and the high number of degrees of freedom (DOF), navigating and path planning for worm-like robots can be difficult [1,2,3]. We have previously shown that even if the robot’s structure is simplified as a series of 2D trapezoids (Figure 1), changing from straight-line locomotion into a turn requires multiple, unique waves that are not periodic. This is in part because the shape of the segments can only be changed within certain bounds as shown in Figure 2 because of the limit of the segment deformation. As a result, both the length traveled and the angle turned for each wave are limited. The compliant modular mesh worm robot with steering (CMMWorm-S)

Methods

Results

Conclusion