For small-scale legged climbing robots, 2D wall-plane traversal with turning is challenging because the required contact force directions and distributions change with motion. For stable climbing, multiple feet must achieve a force-closure contact condition that provides sufficient net forces and moments, rather than relying on simple directional hooking at a single foot. Turning further complicates locomotion because it requires compensating for left–right path-length mismatch without losing fixed foot-tip attachments, demanding complex leg control.

In this project, I conducted foundational studies toward small-scale climbing locomotion, including (1) microspine-based grasping mechanism concepts that enable each foot to attach to the wall beyond simple hooking, and (2) a legged crawling mechanism enabling slipless turning by changing the effective rotation radius of the leg trajectories.

(1) Sheet-metal stamped multi-spine perching mechanism

To increase the probability of successful attachment on rough surfaces, it is beneficial to use many microspines. However, assembling a large number of discrete spines is labor-intensive and difficult. To address this, I fabricated a 2D layer-based laminated mechanism that embeds sheet metal patterned with spine features. The patterned sheet metal was then stamped to form an array of microspines with the desired 3D geometry. This enabled perching mechanism that achieves multi-spine engagement without complex assembly.

(2) Bio-inspired sequential claw–tibial spur grasping mechanism

Inspired by insect feet, where the claw and tibial spur can grip rough substrates, I developed a bio-inspired grasping mechanism that reproduces a sequential engagement: the tibial spur contacts the surface first, followed by claw engagement. The design was inspired by the insect tarsal chain in its stiffness distribution and actuation.

Instead of turning by left–right rotation-count differences (which inherently induces slip), I introduced a transmission that keeps the rotation count the same while changing the effective rotation radius of the left and right leg trajectories, enabling slip-reduced turning. Although the system was not extended to a full climbing robot, I fabricated a minimal two-leg prototype and experimentally demonstrated variable rotation radius.