Magnecko: Design and Control of a Quadrupedal Magnetic Climbing Robot
🤖 AI Summary
To address high-risk industrial inspection in complex 3D ferromagnetic environments—including 90° concave corners, vertical walls, and horizontal overhangs—this paper introduces Magnecko, a quadrupedal magnetic climbing robot. Methodologically, it features a novel insect-inspired 12-DOF biomimetic leg architecture integrating electromagnetically controlled permanent magnet adhesion and high-torque micro-actuators, enabling unified adhesion-locomotion design. A biomimetic leg kinematic model is formulated, and motion planning combines model predictive control (MPC) with real-time gait generation to support full-pose load-bearing climbing and cross-surface continuous locomotion. Experimentally, Magnecko demonstrates stable locomotion on ground, vertical walls, and horizontal overhangs; negotiates multi-angle corner obstacles; and sustains ≥65% of its body weight under arbitrary orientations. This work establishes a reproducible, extensible research platform for legged magnetic climbing robots.
Application Category
📝 Abstract
Climbing robots hold significant promise for applications such as industrial inspection and maintenance, particularly in hazardous or hard-to-reach environments. This paper describes the quadrupedal climbing robot Magnecko, developed with the major goal of providing a research platform for legged climbing locomotion. With its 12 actuated degrees of freedom arranged in an insect-style joint configuration, Magnecko's high manipulability and high range of motion allow it to handle challenging environments like overcoming concave 90 degree corners. A model predictive controller enables Magnecko to crawl on the ground on horizontal overhangs and on vertical walls. Thanks to the custom actuators and the electro-permanent magnets that are used for adhesion on ferrous surfaces, the system is powerful enough to carry additional payloads of at least 65 percent of its own weight in all orientations. The Magnecko platform serves as a foundation for climbing locomotion in complex three-dimensional environments.
Problem
Research questions and friction points this paper is trying to address.
Design a quadrupedal robot for climbing complex 3D environments
Enable locomotion on walls and overhangs using magnetic adhesion
Achieve payload capacity of 65% robot weight in all orientations
Innovation
Methods, ideas, or system contributions that make the work stand out.
Quadrupedal robot with 12 actuated degrees
Model predictive controller for versatile crawling
Electro-permanent magnets enable strong adhesion
🔎 Similar Papers
No similar papers found.
