LEGO: Latent-space Exploration for Geometry-aware Optimization of Humanoid Kinematic Design

2026-04-09

AI Summary

This work proposes a data-driven, automated paradigm for robotic kinematic design that overcomes the limitations of traditional approaches relying on human intuition, which struggle with an intractably large design space and ill-defined task-specific loss functions. The method learns a compact, geometry-preserving latent space from existing mechanical structures, represents joint axes using screw theory, and formulates an optimization objective through motion retargeting and Procrustes analysis based on human motion data. Gradient-free optimization in this latent space jointly synthesizes anthropomorphic upper-limb kinematic structures. Notably, the approach eliminates the need for manually constructing either the design space or task losses, achieving—for the first time—geometry-aware latent-space optimization and demonstrating the effective discovery of novel robotic morphologies directly from existing designs and human motion.

Technology Category

• Intelligent Robots: Learning & Optimization for ROB
• Machine Learning: Learning with Manifolds
• Humans and AI: Human-Aware Planning and Behavior Prediction

Application Category

• Responsible Web: Machine-in-the-loop, human agency and autonomy
• Semantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactions
• User Modeling, Personalization and Recommendation: Fairness-aware retrieval and ranking

Abstract

Designing robot morphologies and kinematics has traditionally relied on human intuition, with little systematic foundation. Motion-design co-optimization offers a promising path toward automation, but two major challenges remain: (i) the vast, unstructured design space and (ii) the difficulty of constructing task-specific loss functions. We propose a new paradigm that minimizes human involvement by (i) learning the design search space from existing mechanical designs, rather than hand-crafting it, and (ii) defining the loss directly from human motion data via motion retargeting and Procrustes analysis. Using screw-theory-based joint axis representation and isometric manifold learning, we construct a compact, geometry-preserving latent space of humanoid upper body designs in which optimization is tractable. We then solve design optimization in this latent space using gradient-free optimization. Our approach establishes a principled framework for data-driven robot design and demonstrates that leveraging existing designs and human motion can effectively guide the automated discovery of novel robot design.

Problem

• robot design
• kinematic optimization
• latent space
• motion retargeting
• design automation

Innovation

• latent-space learning
• geometry-aware optimization
• motion retargeting
• screw theory
• design co-optimization