Learning to Control Self-Assembling Morphologies: A Study of Generalization via Modularity
Contemporary sensorimotor learning approaches typically start with an existing complex agent (e.g., a robotic arm), which they learn to control. In contrast, this paper investigates a modular co-evolution strategy: a collection of primitive agents learns to dynamically self-assemble into composite bodies while also learning to coordinate their behavior to control these bodies. Each primitive agent consists of a limb with a motor attached at one end. Limbs may choose to link up to form collectives. When a limb initiates a link-up action, and there is another limb nearby, the latter is magnetically connected to the 'parent' limb's motor. This forms a new single agent, which may further link with other agents. In this way, complex morphologies can emerge, controlled by a policy whose architecture is in explicit correspondence with the morphology. We evaluate the performance of these dynamic and modular agents in simulated environments. We demonstrate better generalization to test-time changes both in the environment, as well as in the structure of the agent, compared to static and monolithic baselines. Project video and code are available at https://pathak22.github.io/modular-assemblies/
Code (1)
Similar Papers 제목 키워드 기반
MetaMorph: Learning Universal Controllers with Transformers
Multiple domains like vision, natural language, and audio are witnessing tremendous progress by leveraging Transformers for large scale pre-training followed by task specific fine tuning. In contrast, in robotics we prim…
Zero-shot GeneralizationSelf-assembly, Buckling & Density-Invariant Growth of Three-dimensional Vascular Networks
The experimental actualisation of organoids modelling organs from brains to pancreases has revealed that much of the diverse morphologies of organs are emergent properties of simple intercellular "rules" and not the resu…
AnyBody: A Benchmark Suite for Cross-Embodiment Manipulation
Generalizing control policies to novel embodiments remains a fundamental challenge in enabling scalable and transferable learning in robotics. While prior works have explored this in locomotion, a systematic study in the…
Zero-shot GeneralizationHeteroMorpheus: Universal Control Based on Morphological Heterogeneity Modeling
In the field of robotic control, designing individual controllers for each robot leads to high computational costs. Universal control policies, applicable across diverse robot morphologies, promise to mitigate this chall…
DiversityZero-shot GeneralizationBeyond Fixed Morphologies: Learning Graph Policies with Trust Region Compensation in Variable Action Spaces
Trust region-based optimization methods have become foundational reinforcement learning algorithms that offer stability and strong empirical performance in continuous control tasks. Growing interest in scalable and reusa…
Reinforcement LearningContinuous Control