Re-Examining Climbing Robots: Design and Performance of a Lightweight, Low-Cost Robot with a Highly Extendable Limb

Climbing robots are a growing area of interest for tasks that involve vertical mobility in locations that are difficult or dangerous to access for humans. These robots are often designed for inspection, surveillance, or maintenance tasks, but have not been widely deployed due to key limitations with existing designs. Wheeled climbers have little to no adaptability to surface variations, and can generally only climb a single flat, featureless surface with no obstacles. Limbed climbers possess much greater adaptability and can easily step over obstacles using limbs with additional DOFs, but are often heavy, slow, and expensive. These shortcomings are mainly due to these robots' reliance on conventional, ground-based mechanical designs that do not account well for certain fundamental differences between climbing and walking. In this paper, we analyze the previously introduced EEWOC (Extended-reach Enhanced Wheeled Orb for Climbing) robot, which is better suited for climbing. It uses a long-reach tape spring limb to extend and bend over obstacles, while reducing the number of actuators by using a simplified morphology with passive and switchable mechanisms. The result is an adaptable, lightweight, low-cost robot with superior climbing speed and payload capacity compared to conventional systems. This paper outlines the fundamental differences between climbing and ground-based movement, discusses the design problems with current climbing robots and how EEWOC addresses them, and finally compiles performance data from a wide variety of other climbing robots, categorizing and comparing them to better understand the impact of EEWOC's design strategies.