Text2Robot: Evolutionary Robot Design from Text Descriptions

TL;DR

Text2Robot is a framework that rapidly converts textual descriptions into physically manufacturable quadrupedal robots through a combination of text-to-3D modeling and co-optimization, significantly accelerating robot design.

Contribution

This work introduces a novel end-to-end system that transforms user text inputs into functional robot prototypes within a day, integrating generative models with geometric and control optimization.

Findings

Generates initial robot designs within minutes from text.

Produces fully optimized, manufacturable robots in about a day.

Enables rapid prototyping and expands design possibilities with generative models.

Abstract

Robot design has traditionally been costly and labor-intensive. Despite advancements in automated processes, it remains challenging to navigate a vast design space while producing physically manufacturable robots. We introduce Text2Robot, a framework that converts user text specifications and performance preferences into physical quadrupedal robots. Within minutes, Text2Robot can use text-to-3D models to provide strong initializations of diverse morphologies. Within a day, our geometric processing algorithms and body-control co-optimization produce a walking robot by explicitly considering real-world electronics and manufacturability. Text2Robot enables rapid prototyping and opens new opportunities for robot design with generative models.