🤖 AI Summary
Conventional quadruped robot design suffers from high cost, long development cycles, and difficulty in simultaneously satisfying user-defined performance preferences and physical manufacturability constraints. Method: This paper introduces the first text-driven co-optimization framework for morphology and control. Given natural-language performance specifications (e.g., “lightweight, high obstacle traversal, low power consumption”), the framework jointly synthesizes physically realizable 3D morphologies—respecting electronic component placement and manufacturing process constraints—and corresponding controller parameters. Our approach integrates text-to-3D generation, geometric processing, co-evolutionary optimization, and manufacturability-aware search. Contribution/Results: Experiments demonstrate that the system generates diverse initial morphologies within minutes and produces physically functional, stably walking prototypes within 24 hours—reducing design time by orders of magnitude and advancing the practical deployment of generative robot design.
📝 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.