Pressure-free Magnetic Soft Growing Robot with Real-Time Shape Control and Sensing for Biome Sampling

Soft growing robots offer a promising solution for deployment in confined and unstructured environments due to their unique locomotion principles. Although this approach has revolutionized the way robots can navigate lumens of varying diameters and shapes, steering these robots remains challenging-especially at bifurcations or when navigating toward specific targets in open-space environments. While ongoing research continues to improve steering capabilities, the shape of soft growing robots is still largely dictated by their surrounding environment. This study addresses this fundamental limitation by extending the concept of magnetic steering previously introduced in the literature. We propose a new approach that employs an embedded magnetic signature to enable full-body shape deformation. Based on a physical model that combines magneto-elasticity with the eversion growth principle, we also introduce a method for shape sensing and demonstrate the concept of pressure-free growth. The robot’s magnetic signature is defined along the X and Y axes, while local magnetization along the Z axis enables shape sensing. As a proof of concept, the proposed soft growing robot has a diameter of 1.8 mm and can evert over a length of 6 cm. It incorporates two submillimeter 3D Hall-effect sensors for real-time shape sensing, a camera, and a cytology brush that can be deployed from the inner channel for biome sampling. The robot can grow without internal pressure, be magnetically steered in open space to reach complex configurations such as retroflexion, and its shape can be sensed at frequencies up to 500 Hz.