An End-to-End Framework for Optimizing Foot Trajectory and Force in Dry Adhesion Legged Wall-Climbing Robots
Jichun Xiao, Jiawei Nie, Lina Hao, Zhi Li · 2025
Read An End-to-End Framework for Optimizing Foot Trajectory and Force in Dry Adhesion Legged Wall-Climbing Robots by Jichun Xiao, Jiawei Nie, Lina Hao, Zhi Li (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2504-19448.txt.
Introduction
A legged robot that climbs walls on dry adhesive feet lives or dies by each step: peel too hard and the foot tears free; press too softly and it never grips. Jichun Xiao, Jiawei Nie, Lina Hao, and Zhi Li present in 2025 a framework that optimizes the whole foot trajectory and the forces along it, end to end.
Smooth three-part curves describe candidate steps for different foot shapes. A learned model predicts the forces any trajectory will produce, and a multi-objective optimizer, steered by a redundancy-aware strategy, picks the step that best trades off grip, peel force, and vibration. On the team's quadruped climber, the chosen trajectories cut peak detachment force by 28 percent and vibration by 82 percent against standard steps. The authors offer the framework as a general recipe for adhesive climbing robots.
Catalog record
| CREATOR | Jichun Xiao, Jiawei Nie, Lina Hao, Zhi Li |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 61 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2504-19448
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
An End-to-End Framework for Optimizing Foot Trajectory and Force in Dry Adhesion Legged Wall-Climbing Robots Jichun Xiao, Jiawei Nie, Lina Hao, Zhi Li Abstract Foot trajectory planning for dry adhesion legged climbing robots presents challenges, as the phases of foot detachment, swing, and adhesion significantly influence the adhesion and detachment forces essential for stable climbing. To tackle this, an end-to-end foot trajectory and force optimization framework (FTFOF) is proposed, which optimizes foot adhesion and detachment forces through trajectory adjustments. This framework accepts general foot trajectory constraints and user-defined parameters as input, ultimately producing an optimal single foot trajectory. It integrates three-segment C 2 C^{2} continuous Bezier curves, tailored to various foot structures, enabling the generation of effective climbing trajectories. A dilate-based GRU predictive model establishes the relationship between foot trajectories and the corresponding foot forces. Multi-objective optimization algorithms, combined with a redundancy hierarchical strategy, identify the most suitable foot trajectory for specific tasks, thereby ensuring optimal performance across detachment force, adhesion force and vibration amplitude. Experimental validation on the quadruped climbing robot MST-M3F showed that, compared to commonly used trajectories in existing legged climbing robots, the proposed framework achieved reductions in maximum detachment force by 28 %, vibration amplitude by 82 %, which ensures the stable climbing of dry adhesion legged climbing robots. I Introduction A legged wall-climbing robot is a mobile robot with leg-like structures, designed to navigate vertical, inclined, or even inverted surfaces. These robots are used in tasks that pose risks or challenges for humans, such as inspection [ 1 ] [ 2 ] , maintenance [ 3 ] , cleaning [ 4 ] [ 5 ] , and search-and-rescue operations [ 6 ] [ 7 ] . The adhesion methods for legged wall-climbing robots can be classified into magnetic adhesion [ 8 ] [ 9 ] , electrostatic adhesion [ 10 ] [ 11 ] , vacuum adhesion [ 3 ] [ 12 ] , and dry adhesion [ 13 ] [ 14 ] . Compared to other adhesion types, legged robots utilizing dry adhesion offer distinct advantages. They achieve energy-efficient climbing by using materials that replicate biological adhesive mechanisms, enabling strong but easily reversible attachment without the need for complex mechanical systems. Dry adhesion is also characterized by quiet operation, minimal vibration, and leaving no residue, making it well-suited for use in sensitive environments such as laboratories, cleanrooms, and non-destructive testing scenarios. These advantages expand the robot’s capabilities across various inspection and maintenance tasks. As a special type of foot climbing robot, the trajectory of the foot end of the dry adhesion foot climbing robot consists of the support phase, the detachment phase, the swing phase and the adhesion phase trajectory. Among them, the support force of the support phase is completely provided by the adhesion force of the adhesion phase, so the design of the adhesion phase trajectory is the foundation for stable climbing. The task of the detachment phase is to remove the foot from the adherent surface and complete the forward movement in concert with the oscillating phase. However, the detachment process is often accompanied by large detachment forces. Therefore, the design of the detachment phase trajectory aims to minimize the detachment force, reduce dynamic interference during the detachment process, and ensure the stability of the overall climb. Currently, most existing dry-adhesive legged climbing robots use conventional foot trajectory planning methods and rely on manual debugging experience to generate a segmented trajectory for a specific robot. These methods cannot optimize the large adhesion force and the small detachment force required for the stable movement of the foot because they do not take the adhesion force and detachment force of the foot as the optimization objectives, which affects the climbing performance and stability of the robot. Moreover, the constraints of different foot detachment structures are not considered, resulting in obvious limitations of the generated trajectories in terms of adaptability and scalability. In addition, the complex design of traditional methods makes it difficult to efficiently migrate and rapidly deploy them to different types of dry-adhesive legged climbing robots. In order to address the limitations of traditional methods in terms of adaptability and scalability, an innovative Foot Trajectory and Force Optimization Framework (FTFOF) is proposed in this paper. The framework is based on a generalized acquisition platform and can generate optimal foot trajectories according to different task requirements by synergistically optimizing key performance indicators such as adhesion force, detachment force, and foot jitter at the foot. Compared with existing methods, FTFOF not only improves the adaptability and scalability of trajectories but also realizes rapid extension and portability to different types of dry-adherent legged climbing robots by designing user-friendly interfaces. This innovative framework provides core algorithmic support for dry adhesion legged climbing robots to achieve efficient and stable climbing in diverse motion scenarios. The main contributions of this paper are summarized as follows. 1) A foot optimization framework for dry adhesion legged climbing robots, termed the Foot Trajectory and Force Optimization Framework(FTFOF), is proposed. This framework adjusts the adhesion and detachment forces through foot trajectory modulation, aimed at obtaining a foot trajectory that is optimal for the given task. 2) A redundancy hierarchical strategy(RHS) has been proposed for finding the optimal solution within the Pareto front that meets the current task requirements by prioritizing multiple objectives and retaining the dominant solution using redundancy factors at each level. 3) A mobile generalized data acquisition platform has been constructed, which can acquire foot force data from different planes and different materials surfaces, and can form a corresponding sample set for different scenarios. II Legged Climbing Robot Foot Optimization Framework This paper proposes an optimization framework for foot trajectories in dry adhesive legged climbing robots, termed the Foot Trajectory and Force Optimization Framework(FTFOF). The framework adjusts the foot adhesion and detachment forces through modulation of the foot trajectory to achieve optimal foot trajectories for the given task. The FTFOF proposed in this paper consists of three main modules: the foot climbable trajectory set generation algorithm module, the generalized data acquisition and offline training module, and the multi-objective optimization algorithm module, and the specific FTFOF algorithm framework is shown in Fig. 1 . Through the coupling between the three modules, the FTFOF algorithm is based on the multi-objective optimization algorithm, which can regulate the parameters of the foot adhesion force, foot detachment force, foot jitter and so on through the basic strategy function formulated in this paper, to obtain the optimal single foot trajectory that meets the current task requirements, which effectively improves the climbing stability of the dry-adhesive foot climbing robot. Fig. 1: Legged Climbing Robot Foot Optimization Framework. II- 1 Foot Climbable Trajectory Set Generation Algorithm The Climbable Foot Trajectory Generation Module is designed to produce climbable foot trajectories suitable for various foot structures in dry adhesive legged climbing robots. The generation of these trajectories involves a two-step process. In the first step, a generalized foot trajectory for quadruped climbing robots is fitted using a three-segment C^{2} continuous Bezier curve. In the second step, specific customized constraints are applied based on the unique foot structures of different climbing robots. The final output is a climbable foot trajectory that incorporates these specific customized constraints, including phases for detachment, adhesion,