The Soft Robotics Embodied Intelligence Challenge

Challenge Overview

For AI to be meaningfully integrated into robotics, we must broaden our understanding of what constitutes the "brain" of a robot. Discussions of AI in robotics typically centre around computation in a classical sense - processors, control algorithms, and data-driven decision-making. However, in soft robotics, intelligence is distributed across the entire system, where morphology, material properties, and environmental interactions play fundamental roles.

In this competition, the goal is to design a low cost, bio-inspired, robotic system that exemplifies embodied intelligence through locomotion. Participants are encouraged to draw inspiration from advancements in the soft robotics field, and explore how nature-inspired designs leverage body-environment interactions to achieve efficient locomotion, whether by crawling, undulating, inching, rolling, or employing novel forms of movement.

The challenge invites participants to study how organisms exploit compliance, distributed actuation, and material properties and translate these into an operational robot. The design should not only be functional, but should also highlight the potential of soft robotics to navigate and adapt to complex environments. Additional points will be awarded for innovative approaches that showcase “spectacular” behaviours arising from the interplay of materials, structure, and control.