Robot Tool Design from Scratch via Behavior-Aware Hierarchical Optimization

2026-09-28

AI Summary

This work addresses the limitation of existing tool design approaches constrained by predefined structures, which hinder joint optimization driven by physical objectives. We propose HOT, a hierarchical optimization framework that for the first time incorporates tool structure into a closed-loop physical optimization process. The upper level employs the BASS algorithm to search discrete structures, while the lower level evaluates behavioral feedback via milestone-based physical simulations, using behavioral evidence to guide efficient exploration and enable the co-evolution of structure, shape, and action. Experiments demonstrate that this method discovers functional structures with minimal search effort, significantly reducing task loss while maintaining high success rates. Furthermore, 3D-printed prototypes successfully executed all tasks on real robots, validating the effectiveness of our zero-shot, task-driven tool design paradigm.

Abstract

The ability to design a tool for a task marks a level of intelligence beyond merely understanding, selecting, or using one. Existing methods for robotic tool design typically optimize a tool's continuous shape and action within a structure that is prescribed or generated beforehand, so the structure itself stays outside the physical optimization loop. We study task-driven tool design from scratch, where tool structure, shape, and action are all derived from the desired physical outcome. Here we show that the three elements can be designed jointly by HOT, a hierarchical optimization whose upper level searches over discrete tool structures with BASS, while lower-level physical optimization evaluates their task behavior and returns milestone progress as behavioral evidence for the search, ultimately providing jointly optimized shape and action. On four tool-use tasks with distinct physical functions, HOT discovers functional structures after evaluating only a small fraction of search spaces containing up to 56 million structures, and the subsequent refinement of their geometry lowers the task loss on all tasks while preserving success, through deformations that are functionally interpretable. Once 3D printed, the tools accomplish all tasks on a real robot with the actions found in simulation. Designing tools from required physical effects, rather than a catalog of known tools, is a step toward the open-ended tool making seen in humans and animals.

Problem

• robot tool design
• task-driven design
• hierarchical optimization
• structure search
• tool-use

Innovation

• Hierarchical Optimization
• Behavior-Aware Search
• Tool Design from Scratch
• Discrete-Continuous Co-optimization
• Sim-to-Real Transfer