From robotic active matter to a "Large Matter Model"

Nonreciprocal interactions can generate collective behaviors unavailable to conventional materials, including sustained waves, transverse transport, and odd mechanical responses. We developed the Magnetomechanically Augmented Spinning Robotic (MASBot) platform to explore this physics in a controllable many-body system. By tuning particle-level interactions, the collective transitions between solid-like, fluid-like, and gas-like phases, providing a programmable platform for connecting microscopic rules to macroscopic organization and function. [ preprint ]

​

Matter that learns: Our next question is whether these microscopic rules can be learned and generalized. We seek a shared local policy that runs independently on every particle and produces collective motion, reconfiguration, transport, and adaptation. Because every particle has the same sensing and action space, different robotic bodies can be composed by changing their number, arrangement, and connectivity. This creates a physical setting to test whether one policy can generalize across scale, morphology, environment, task, and changing physics.

​

The long-term vision is a "Large Matter Model": a foundation model that generates diverse physical structures and functions from a common vocabulary of intelligent particles. Just as large language models learn transferable rules over linguistic tokens, a Large Matter Model would learn transferable interaction rules over physical building blocks. By combining decentralized learning with programmable matter, we aim to create materials that can sense, adapt, and reorganize their own function. As the particles become smaller and more numerous, the distinction between a robotic swarm and an intelligent material begins to disappear.