Litcius/Paper detail

Soft Robotic Oscillators With Strain-Based Coordination

Steven Ceron, Marta An Kimmel, Alexandra Q. Nilles, Kirstin Petersen

2021IEEE Robotics and Automation Letters19 citationsDOI

Abstract

Many modular robots and active matter platforms are inspired by natural aggregates of single-celled organisms that exhibit complex emergent behaviors far beyond the capability and range of individuals. These behaviors stem primarily from short range chemical and mechanical interactions. Past work has focused largely on wireless communication akin to chemicals, however, physical interactions require a shift from rigid to soft, durable robots capable of inducing and measuring strain. Here, we present a platform to support such studies. The hardware consists of stand-alone, soft, pneumatic robots capable of radial expansion and contraction. The robots are cheap and fast to produce; they have 3 directional strain sensors, and 6 magnet pairs to loosely couple with their neighbors. We characterize force profiles, sensors, elastic modulus, magnetic interaction, as well as exploratory aggregate motions. Finally, we demonstrate their ability to synchronize, locomote, and fracture in a complimentary quasi-static simulator, with a coupled oscillator model, and discuss evaluation metrics. We hope that this platform will serve to further insights on how simple physical interactions between locally informed agents may lead to complex emergent behaviors.

Topics & Concepts

RobotModular designComputer scienceSoft roboticsWirelessSimulationHuman–computer interactionArtificial intelligenceTelecommunicationsOperating systemModular Robots and Swarm IntelligenceMicro and Nano RoboticsAdvanced Materials and Mechanics