This study aims to develop a novel snake-inspired soft robot prototype designed for efficient crawling within complex and confined pipelines. As a preliminary proof-of-concept, this study evaluates the fundamental locomotion mechanisms and feasibility of the proposed design, aiming to bridge laboratory research with a broad spectrum of future applications, including but not limited to clinical minimally invasive interventions and industrial inspection.
The proposed robot integrates a flexible solenoid coil, a moving permanent magnet, a fixed head magnet and a sawtooth-textured silicone sleeve that generates directional friction against the pipe wall. During each voltage pulse, the internal magnet is driven toward the head to elongate the compliant body, while magnetic repulsion and elastic recovery reset the actuator after the pulse. Locomotion performance is experimentally evaluated in straight and bent PVC tubes at equivalent drive levels of 6–12V. In addition, a lumped-parameter dynamic model is identified from synchronized displacement and current measurements to capture the robot’s axial dynamics.
The robot demonstrates reliable forward crawling in straight pipes and successfully traverses bent PVC tubes with bend angles up to 45°. In a 0.5 m straight tube, the prototype achieves increasing locomotion speed with increasing drive level, reaching a maximum of 0.24 body lengths per second at 12V. The identified lumped-parameter model reproduces the main measured acceleration transients with good agreement, showing that the dominant dynamics of the soft in-pipe crawler can be represented using a compact experimentally fitted framework.
This study presents a soft in-pipe crawling robot that combines embedded electromagnetic actuation with frictional rectification in a compact compliant structure. Unlike many soft crawling systems that rely on pneumatic supply or external magnetic field generators, the proposed design uses an internal coil–magnet arrangement and a textured silicone sleeve to achieve rectilinear locomotion in confined channels. The work also contributes an experimentally identified dynamic model that links electrical input to locomotion behavior and supports future design optimization and control development.
