Scaling hard vertical surfaces with compliant microspine arrays

Alan T. Asbeck, Sangbae Kim, Mark R. Cutkosky, William R. Provancher, Michele Lanzetta

Abstract

A new approach for climbing hard vertical surfaces has been developed that allows a robot to scale concrete, stucco, brick and masonry walls without using suction or adhesives. The approach is inspired by the mechanisms observed in some climbing insects and spiders and involves arrays of microspines that catch on surface asperities. The arrays are located on the toes of the robot and consist of a tuned, multi-link compliant suspension. In this paper we discuss the fundamental issues of spine allometric scaling versus surface roughness and the suspension needed to maximize the probability that each spine will find a useable surface irregularity and to distribute climbing tensile and shear loads among many spines. The principles are demonstrated with a new climbing robot that can scale a wide range of exterior walls.

Introduction

Climbing robots often struggle with "hard" vertical surfaces such as concrete, stucco, and masonry. These materials are ubiquitous in human environments but present significant challenges: they are too hard for traditional claws to penetrate, yet too rough and dusty for suction cups or chemical adhesives to maintain a reliable grip. To address this, researchers have looked to nature, specifically insects and spiders, which utilize arrays of tiny spines to engage with microscopic surface irregularities known as asperities.

Modeling Spine-Surface Interaction

The core of the proposed approach lies in understanding how a small spine "perches" on a rough surface. The researchers developed a 2D geometric model to simulate this interaction. Central to this model is the concept of a "traced surface"—the path described by the tip of a spine as it is dragged across a surface profile.

Geometric model of a microspine interacting with surface asperities, illustrating the 'traced surface' and approach vector.

By analyzing these traced surfaces, the probability of a spine finding a usable "hook" can be calculated based on its tip radius ($r_s$) and the approach angle ($\theta_a$). This modeling revealed a fundamental trade-off: while smaller spines can find more potential attachment points on a fractal-like surface, the maximum load an individual contact can sustain decreases significantly as the spine tip radius shrinks.

Scaling Analysis and Material Limits

The effectiveness of a microspine array depends on the scaling relationship between the spine size and the surface roughness. The researchers analyzed three primary failure modes: plastic bending of the steel spine, elastic slipping, and the brittle failure of the surface asperity itself.

Scaling of asperity density and spine failure force across different materials and length scales.

As shown in the scaling analysis, as the length scale decreases, the density of available asperities increases, but the force each can support drops. On materials like sandpaper or rough concrete, the limit to climbing is often the structural integrity of the wall's surface rather than the strength of the robot's spines. To compensate for the low load capacity of individual small spines, a large number of them must work in parallel.

Hardware Design: SpinyBotII

To implement these findings, the team developed SpinyBotII, a 0.4 kg climbing robot. The critical innovation in its design is a multi-link compliant suspension system for the spine arrays. Using Shape Deposition Manufacturing (SDM), the researchers created "toes" that integrate rigid and flexible materials. This compliance allows each of the 20 spines per foot to independently "search" the surface and engage with an asperity without being hindered by its neighbors.

The robot utilizes an alternating tripod gait and a long tail to counteract the pitching moment that would otherwise pull its upper body away from the wall.

Results and Discussion

SpinyBotII successfully scaled a variety of challenging terrains, including brick, concrete, stucco, and dressed stone, at speeds of approximately 2.3 cm/s. The optimal configuration utilized spines with a 15 $\mu$m tip radius, with each contact capable of sustaining 1–2 N of force.

While successful, the current 2D modeling does not fully account for 3D surface complexities where spines might be steered into or away from grooves. Future work will focus on 3D modeling, increasing robot agility, and exploring hybrid attachment systems that combine microspines with gecko-inspired adhesives for performance on both rough and smooth surfaces.

Summary updated

Abstract

A new approach for climbing hard vertical surfaces has been developed that allows a robot to scale concrete, stucco, brick and masonry walls without using suction or adhesives. The approach is inspired by the mechanisms observed in some climbing insects and spiders and involves arrays of microspines that catch on surface asperities. The arrays are located on the toes of the robot and consist of a tuned, multi-link compliant suspension. In this paper we discuss the fundamental issues of spine allometric scaling versus surface roughness and the suspension needed to maximize the probability that each spine will find a useable surface irregularity and to distribute climbing tensile and shear loads among many spines. The principles are demonstrated with a new climbing robot that can scale a wide range of exterior walls.
BibTeX
@article{asbeck2006scaling,
  author = {Alan T. Asbeck and Sangbae Kim and Mark R. Cutkosky and William R. Provancher and Michele Lanzetta},
  title = {{Scaling hard vertical surfaces with compliant microspine arrays}},
  journal = {The International Journal of Robotics Research},
  year = {2006},
  doi = {10.1177/0278364906072511},
}
Figure