The emerging field of morphological computation seeks to understand how mechanical complexity in living systems can be advantageous, for instance by reducing the cost of control. In this paper we explore the phenomenon of morphological computation in tensegrities -unique structures with a high strength to weight ratio, resilience, and an ability to change shape. These features have great value as a robotics platform, but also make tensegrities difficult to control via conventional techniques. We describe a novel approach to the control of tensegrity robots which, rather than suppressing complex dynamics, exploits them in order to achieve locomotion. Our robots are physically embodied (rather than simulated), evolvable, and locomote at higher speeds (relative to body size) and with fewer actuators than those controlled by more conventional approaches.

Due to their high strength-to-weight ratio, robustness and deformability, tensegrity structures are an appealing platform for the emerging field of soft robotics, with applications ranging from search-and-rescue to field-deployable structures. Unfortunately, these properties also make tensegrities challenging to control through conventional means. In this paper we describe a novel means of vibration-based tensegrity actuation which allows for the manual control of a physical tensegrity robot in the plane as well as state-machine based target following. We demonstrate the evolution of effective gaits using only physical evaluations of the robot, and further demonstrate how a combination of the state-machine with the hill climber allows for the hands-off automation of the evolutionary process. We conclude with a description of how this can lead to a bootstrapping effect, with the potential to significantly accelerate and automate the physical evolution of our tensegrity robot.

Conventionally control can be achieved by attempting to simplify complex dynamics. The field of morphological computation explores how mechanical complexity can be advantageous. In this paper we demonstrate morphological computation in tensegrity robots. We present a novel approach to tensegrity actuation and explore the capabilities of our self-evolving system. Methods of finding desirable gaits through both hand selection and evolution are described and the effectiveness of the system is demonstrated by our robot's ability to pursue a moving target. We conclude with a discussion of a bootstrapped system with the potential of significantly reducing evolution time and need for user presence.

No abstract