Lateral Undulation of the Bendable Body of a Gecko-Inspired Robot for Energy-Efficient Inclined Surface Climbing
Worasuchad Haomachai,Donghao Shao,Wei Wang,Aihong Ji,Zhendong Dai,Poramate Manoonpong
DOI: https://doi.org/10.1109/lra.2021.3101519
IF: 5.2
2021-10-01
IEEE Robotics and Automation Letters
Abstract:Sprawling posture animals with their bendable spine, such as salamanders, and geckos, can perform agile and versatile locomotion including walking, swimming, and climbing. Therefore, several roboticists have used them as templates for robot designs to investigate and generate efficient locomotion. Typically, walking and/or swimming abilities are realized by salamander-inspired robots with a bendable body, whereas climbing ability is achieved on gecko-inspired robots with an over-simplified fixed body. In this study, we propose optimal bendable body design with three degrees of freedom (DOFs). Its implementation on a sprawling posture robot is inspired by geckos for climbing enhancement. The robot leg and body movements are coordinated and driven by central pattern generator (CPG)-based neural control. As a consequence, the robot can climb using a combination of trot gait and lateral undulation of the bendable body with a C-shaped standing wave. Through the real robot experiments on a 3D force measuring platform, we demonstrate that, due to the dynamics of the bendable body movement, the robot can gain higher mediolateral ($F_x$) ground reaction forces (GRFs) at its front legs as well as anteriorposterior ($F_y$) GRFs at its hind legs to increase the bending angular momentum ($L_A{}_M$). This results in 52 and 54 reduced energy consumptions during climbing on steeper inclined solid and soft surfaces, respectively, compared to climbing with a fixed body. To this end, the study provides a basis for developing sprawling posture robots with a bendable body and neural control for energy-efficient inclined surface climbing with a possible extension towards agile and versatile locomotion, such as sprawling posture animals.
robotics