MetaMorph: Learning Universal Controllers with Transformers
This addresses the challenge of impractical per-morphology training in robotics, enabling flexible and efficient control for modular systems.
The paper tackles the problem of training controllers for the exponentially large number of possible modular robot morphologies by proposing MetaMorph, a Transformer-based approach that learns a universal controller, achieving zero-shot generalization to unseen morphologies and sample-efficient transfer to new tasks.
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 primarily train a single robot for a single task. However, modular robot systems now allow for the flexible combination of general-purpose building blocks into task optimized morphologies. However, given the exponentially large number of possible robot morphologies, training a controller for each new design is impractical. In this work, we propose MetaMorph, a Transformer based approach to learn a universal controller over a modular robot design space. MetaMorph is based on the insight that robot morphology is just another modality on which we can condition the output of a Transformer. Through extensive experiments we demonstrate that large scale pre-training on a variety of robot morphologies results in policies with combinatorial generalization capabilities, including zero shot generalization to unseen robot morphologies. We further demonstrate that our pre-trained policy can be used for sample-efficient transfer to completely new robot morphologies and tasks.