← Back ICRA 2026

PIMBS: Efficient Body Schema Learning for Musculoskeletal Humanoids with Physics-Informed Neural Networks

Kento Kawaharazuka, Takahiro Hattori, Keita Yoneda, Kei Okada

PDF

AI summary

Key figure (auto-extracted from paper)

Embedding physical torque-tension constraints into neural network training enables accurate body schema learning for complex musculoskeletal robots using only a handful of real-world data points.

Musculoskeletal Humanoids Body Schema Learning Physics-Informed Neural Networks Data-Efficient Learning Robot Calibration Variable Stiffness Control

Problem

Geometric models of musculoskeletal humanoids often deviate from reality, making body schema learning necessary. However, collecting sufficient data from actual robots is labor-intensive, hindering accurate learning when data is scarce.

Approach

The authors introduce PIMBS, which integrates physics-based constraints and boundary conditions directly into the neural network's loss function. This allows the model to leverage differential information and physical laws to learn muscle-joint relationships efficiently from minimal data.

Key results

60–70% error reduction with only 3–5 training samples in simulation
Combined physics and boundary losses maximize accuracy across all data regimes
Successful real-world validation on a 5-DOF 10-muscle musculoskeletal arm
Physics constraints improve extrapolation accuracy at data-sparse joint extremes

Why it matters

Provides a practical, data-efficient calibration framework for complex bio-inspired robots, accelerating their real-world deployment and control.

Abstract

Musculoskeletal humanoids are robots that closely mimic the human musculoskeletal system, offering various advantages such as variable stiffness control, redundancy, and flexibility. However, their body structure is complex, and muscle paths often significantly deviate from geometric models. To address this, numerous studies have been conducted to learn body schema, particularly the relationships among joint angles, muscle tension, and muscle length. These studies typically rely solely on data collected from the actual robot, but this data collection process is labor-intensive, and learning becomes difficult when the amount of data is limited. Therefore, in this study, we propose a method that applies the concept of Physics- Informed Neural Networks (PINNs) to the learning of body schema in musculoskeletal humanoids, enabling high-accuracy learning even with a small amount of data. By utilizing not only data obtained from the actual robot but also the physical laws governing the relationship between torque and muscle tension under the assumption of correct joint structure, more efficient learning becomes possible. We apply the proposed method to both simulation and an actual musculoskeletal humanoid and discuss its effectiveness and characteristics.

Index terms

Tendon/Wire Mechanism Learning from Experience Human and Humanoid Motion Analysis and Synthesis