Evolution of Locomotion on a Physical Tensegrity Robot

Khazanov, Mark; Jocque, Julian; Rieffel, John

doi:10.7551/978-0-262-32621-6-ch039

Cited by 24 publications

(14 citation statements)

References 16 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Koizumi et al use a tethered icosahedron robot with pneumatic actuators to analyze base patterns for low-energy rolling [17]. Rieffel et al use an icosahedron robot with vibrational motors analyzing forward motion with vibration [27,26,15]. Suitable for rolling locomotion, ReCTeR is the closest to the ideal icosahedron robot [9].…”

Section: Background and Previous Workmentioning

confidence: 97%

Learning Tensegrity Locomotion Using Open-Loop Control Signals and Coevolutionary Algorithms

et al. 2015

View full text Add to dashboard Cite

Soft robots offer many advantages over traditional rigid robots. However, soft robots can be difficult to control with standard control methods. Fortunately, evolutionary algorithms can offer an elegant solution to this problem. Instead of creating controls to handle the intricate dynamics of these robots, we can simply evolve the controls using a simulation to provide an evaluation function. In this article, we show how such a control paradigm can be applied to an emerging field within soft robotics: robots based on tensegrity structures. We take the model of the Spherical Underactuated Planetary Exploration Robot ball (SUPERball), an icosahedron tensegrity robot under production at NASA Ames Research Center, develop a rolling locomotion algorithm, and study the learned behavior using an accurate model of the SUPERball simulated in the NASA Tensegrity Robotics Toolkit. We first present the historical-average fitness-shaping algorithm for coevolutionary algorithms to speed up learning while favoring robustness over optimality. Second, we use a distributed control approach by coevolving open-loop control signals for each controller. Being simple and distributed, open-loop controllers can be readily implemented on SUPERball hardware without the need for sensor information or precise coordination. We analyze signals of different complexities and frequencies. Among the learned policies, we take one of the best and use it to analyze different aspects of the rolling gait, such as lengths, tensions, and energy consumption. We also discuss the correlation between the signals controlling different parts of the tensegrity robot.

show abstract

Section: Background and Previous Workmentioning

confidence: 97%

Learning Tensegrity Locomotion Using Open-Loop Control Signals and Coevolutionary Algorithms

et al. 2015

View full text Add to dashboard Cite

show abstract

“…Online controller learning is relatively easy to implement with physical experimentation, as the morphology remains fixed, and it has been applied successfully for locomotion [53,85] and a soft fish robot [123]. Other studies have co-optimized control and morphology in the real world.…”

Section: Scaling Up Physical Experimentationmentioning

confidence: 99%

Reality-Assisted Evolution of Soft Robots through Large-Scale Physical Experimentation: A Review

et al. 2020

View full text Add to dashboard Cite

We introduce the framework of reality-assisted evolution to summarize a growing trend towards combining model-based and model-free approaches to improve the design of physically embodied soft robots. In silico, data-driven models build, adapt, and improve representations of the target system using real-world experimental data. By simulating huge numbers of virtual robots using these data-driven models, optimization algorithms can illuminate multiple design candidates for transference to the real world. In reality, large-scale physical experimentation facilitates the fabrication, testing, and analysis of multiple candidate designs. Automated assembly and reconfigurable modular systems enable significantly higher numbers of real-world design evaluations than previously possible. Large volumes of ground-truth data gathered via physical experimentation can be returned to the virtual environment to improve data-driven models and guide optimization. Grounding the design process in physical experimentation ensures that the complexity of virtual robot designs does not outpace the model limitations or available fabrication technologies. We outline key developments in the design of physically embodied soft robots in the framework of reality-assisted evolution.

show abstract

“…A six-rod tensegrity structure that resembles an icosahedron is the simplest three-dimensional (3D) spherical tensegrity structure, and several robots have been introduced in the literature based on this structure. 20,21,[23][24][25][26][27][28][29][30] Some research on these robots developed locomotion strategies from trial-and-error hardware experiments, 25,26,31,32 which can be time consuming and are not scalable to the robots with a larger number of members. To overcome this difficulty, other researchers developed locomotion strategies first in simulation, usually in conjunction with learning algorithms, to manage the inherent structural complexity of tensegrities, and then validated them with hardware robots.…”

Section: Introductionmentioning

confidence: 99%

Rolling Locomotion of Cable-Driven Soft Spherical Tensegrity Robots

Kim

Agogino

2020

Soft Robotics

View full text Add to dashboard Cite

Soft spherical tensegrity robots are novel steerable mobile robotic platforms that are compliant, lightweight, and robust. The geometry of these robots is suitable for rolling locomotion, and they achieve this motion by properly deforming their structures using carefully chosen actuation strategies. The objective of this work is to consolidate and add to our research to date on methods for realizing rolling locomotion of spherical tensegrity robots. To predict the deformation of tensegrity structures when their member forces are varied, we introduce a modified version of the dynamic relaxation technique and apply it to our tensegrity robots. In addition, we present two techniques to find desirable deformations and actuation strategies that would result in robust rolling locomotion of the robots. The first one relies on the greedy search that can quickly find solutions, and the second one uses a multigeneration Monte Carlo method that can find suboptimal solutions with a higher quality. The methods are illustrated and validated both in simulation and with our hardware robots, which show that our methods are viable means of realizing robust and steerable rolling locomotion of spherical tensegrity robots.

show abstract

Evolution of Locomotion on a Physical Tensegrity Robot

Cited by 24 publications

References 16 publications

Learning Tensegrity Locomotion Using Open-Loop Control Signals and Coevolutionary Algorithms

Learning Tensegrity Locomotion Using Open-Loop Control Signals and Coevolutionary Algorithms

Reality-Assisted Evolution of Soft Robots through Large-Scale Physical Experimentation: A Review

Rolling Locomotion of Cable-Driven Soft Spherical Tensegrity Robots

Contact Info

Product

Resources

About