A Review of Research on Crawling Soft Robots: Bionic Mechanisms and Actuation Motion.

Crawling soft robots have attracted widespread attention due to their high adaptability to unstructured environments. However, existing research often focuses on understanding the formation mechanism of their motion capabilities from the perspective of actuation methods or control strategies, resulting in fragmented design logic and a lack of a possible framework. Through comparative analysis across animal, plant, and microbial systems, this paper points out that crawling behavior in different biological systems largely depends on the synergistic effect of morphological deformation and interface friction. Many studies have shown that rectifying periodic, reversible deformation processes into directional net displacement plays a crucial role. Building on this, this paper further analyzes the roles of various actuation technologies in crawling systems, emphasizing that actuation primarily undertakes deformation triggering and modulation functions, and its impact on motion performance is highly dependent on the coupling method with morphological structure and interface conditions. Regarding control and learning methods, this paper discusses the key role of morphological and interface design in reducing control dimensionality and improving system robustness from the perspective of embodied intelligence, pointing out that control strategies are more about compensating for and optimizing the structure-generated motion capabilities. The motion capability of soft crawling robots is not entirely determined by a single actuation performance, but rather stems from the synergistic rectification of periodic deformation by morphological structure, actuation timing, and interface interactions. Finally, this paper summarizes and proposes several future-oriented design languages and research paradigms, providing a unified reference framework for understanding the mechanisms and engineering design of crawling soft robots.