ROCR: Dynamic vertical wall climbing with a pendular two-link mass-shifting robot

Abstract

We present a novel bio-inspired dynamic climbing robot, with a recursive name: ROCR is an Oscillating Climbing Robot. ROCR is a pendular two-link, serial chain robot that utilizes alternating hand-holds and an actuated tail to propel itself upward in a climbing style based on observation of human climbers and brachiating gibbons. ROCR's bio-inspired oscillating climbing strategy is simple and efficient. This robot is intended for autonomous surveillance and inspection on sheer vertical surfaces. Potential locomotion gait strategies were investigated in simulation using Working Model 2D. Strategy comparisons were drawn on a basis of climbing rate and energy efficiency. Simulation results show the feasibility and relative merits of these climbing gait strategies and compare them to the potential energy gained while climbing. A first generation ROCR design, having physical parameters based on simulation analysis, that utilizes actuated permanent magnet grippers and a geared-motor actuated tail was fabricated. Magnetic grippers are used for this initial prototype to facilitate investigating the robots climbing efficiency. Several climbing gaits which showed promise in simulation were implemented in the prototype. ROCR is a lightweight, flexible, self-contained robot with onboard microcontroller, tilt and acceleration sensing, power supply and wireless communication. We conclude with plans to evolve the simulation and design of the robot and plans for testing the prototype.

Hovering of MAV by using magnetic adhesion and winch mechanisms

“…In addition, we have added a new function by which an MAV approaches a target object from above by using a winch mechanism. Many approaches have been proposed for adhering to walls and ceilings: negative pressure [4], [5], magnetic adhesion [6], [7], micro spines [8], electrical adhesion [9], [10], adhesive pad [11], [12], and biomimetics [13], [14]. Our target environment is industrial plants and damaged buildings.…”

“…In Sections III and IV, we describe mechanisms mounted on an MAV. In Section V, we show TABLE I COMPARISON OF ADHESION METHODS (++:FAVORABLE, +:LITTLE FAVORABLE, -:SLIGHTLY UNFAVORABLE, −−:UNFAVORABLE) adhesion low power adhesion force continuous less danger attaching to method consumption for weight using to surroundings diverse substance negative pressure [4], [5] -++ ++ + ++ magnetic adhesion [6], [7] ++ ++ ++ ++ −− micro spines [8] ++ + ++ -+ electrical adhesion [9], [10] + -+ -++ adhesive pad [11], [12] ++ + -+ -biomimetics [13], [14] ++ --+ -how an MAV can adhere to a ceiling, lower itself on a tether, and observe objects.…”

On the Comparative Analysis of Locomotory Systems with Vertical Travel

“…While specific resistance has been compared for a variety of ground vehicles-an exhaustive comparison at time of publication is in [8]-no similar comparison has been performed for climbing robots. With dynamic and efficient climbers [3,5,10], as well as various other robots that climb on a variety of surfaces at differing speeds [13,1,9], we believe the lack of a formal comparison to be something that needs addressing in the climbing robot community. This paper outlines a methodology by which specific resistance can be applied to climbing robots-even those that only climb on sloped surfaces rather than vertical, such as in [2,5]-and compared to other climbing robots as well as to the growing list of ground robots covered in work such as [8].…”

Dynamic climbing of near-vertical smooth surfaces

“…DynoClimber is capable of climbing at 1.5 body-lengths per second (66 cm/s) and is a *These authors contributed equally to this work. This work is supported by the National Science Foundation by Grant EEC-0832819 and Grant 0856789 P. Birkmeyer and R. S. Fearing are with the Department of Electrical Engineering and Computer Science, University of California, Berkeley{paulb ronf} at eecs.berkeley.edu A. G. Gillies is with the Department of Mechanical Engineering, University of California, Berkeley -andrew.gillies at berkeley.edu novel adaptation of a biologically-derived climbing template [14]; ROCR uses unique tail-swinging dynamics to achieve climbing [15]. However, both robots chose simple climbing surfaces in order to not impede the development and study of their lateral-plane dynamics.…”