Abstract
Elongate, limbless robots are uniquely suited for navigating cluttered environments, yet executing sharp, efficient turns remains a significant challenge. This paper presents a robust turning template derived from the biological "omega turn" of the nematode C. elegans. By modeling the turn as a superposition of two traveling waves, we demonstrate a controller that allows both snake-like and multi-legged robots to achieve high-angle, compact turns in complex terrains, including granular media and outdoor rock piles.
Introduction
Robotic snakes and other elongate systems are ideal for search-and-rescue or industrial inspection in tight spaces. However, while forward locomotion is well-understood, effective turning is difficult to implement. Traditional strategies, such as adding a constant offset to a body wave, often result in wide, sweeping movements that lead to collisions in confined areas. The goal of this research is to develop a turning strategy that is both agile (large turn angle) and compact (small footprint) while remaining robust to unpredictable environmental interactions.
Biological Inspiration and Methodology
The researchers employed a comparative theory-biology approach, beginning with the microscopic worm C. elegans. By tracking the worm's movements and applying Principle Component Analysis (PCA), they decomposed the complex "omega turn"—where the organism's head nearly touches its tail—into fundamental mathematical components.
The Wave Superposition Template
The core discovery was that the omega turn can be prescribed by superimposing two distinct traveling waves: a "forward wave" and a "turning wave" with different spatial frequencies. Using geometric mechanics, the team calculated "height functions" to optimize the timing and amplitude of these waves, maximizing the turn angle while preventing body self-collision.
To handle real-world obstacles, the researchers implemented a compliant control framework. This allows the robot to adjust wave amplitudes based on sensed joint torques, effectively using physical contact with the environment to facilitate the turn.
Experimental Results
The template was tested on a 16-joint snake robot and a 10-legged elongate robot across diverse environments.
- Superior Agility: On flat ground, the omega turn achieved an average angular displacement of 105.7° per cycle, vastly outperforming traditional offset turns (61.6°) and geometric turns (9.0°).
- Compactness: The omega turn demonstrated the smallest "swept area" per degree of turn, making it the most space-efficient strategy for cluttered environments.
- Environmental Robustness: In field tests on random rock piles, the omega turn maintained a robust 100.1° per cycle. In granular media (simulating sand), it produced ~78.5° of turn, whereas other common strategies failed to produce significant reorientation.
Discussion and Future Work
The universality of this template is a key finding; the same wave superposition enabled a multi-legged robot to turn effectively, suggesting that body-driven turning is a fundamental strategy for any elongate system. Furthermore, mechanical intelligence—leveraging physical interactions via compliant control—improved turning performance by approximately 30%.
Future research will focus on developing autonomous transitions between forward crawling and the omega turn, as well as integrating the template into high-level path planners for fully autonomous missions. Extending these 2D motions into 3D "volumetric" turns could further enhance maneuverability in complex vertical structures.
Summary updated