MOB-Net: Limb-modularized Uncertainty Torque Learning of Humanoids for Sensorless External Torque Estimation

2024-02-17

Int. J. Robotics Res.

AI Summary

To address the poor robustness and large estimation bias of conventional momentum observers (MOBs) for floating-base humanoid robots—stemming from modeling inaccuracies and friction in the absence of external torque sensors—this paper proposes a modular uncertainty torque learning architecture. The approach decouples physics-based MOB estimation from data-driven residual calibration, fusing IMU and joint encoder measurements. It incorporates limb-modular neural networks, uncertainty torque residual learning, and floating-base kinematic constraints to enable end-to-end joint optimization. The method significantly enhances generalization to unseen operational conditions: simulation and real-world experiments on the KHR-3HV platform demonstrate a 62% reduction in torque estimation error. Furthermore, it enables millisecond-level collision detection, contact force feedback control, and autonomous safety responses.

Technology Category

• Intelligent Robots: State Estimation
• Machine Learning: Mixture of Experts (MoE)
• Humans and AI: Human-Aware Planning and Behavior Prediction

Application Category

• Responsible Web: Machine-in-the-loop, human agency and autonomy
• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications
• User Modeling, Personalization and Recommendation: On-Device user modeling, personalization, and recommendation

Abstract

Momentum observer (MOB) can estimate external joint torque without requiring additional sensors, such as force/torque or joint torque sensors. However, the estimation performance of MOB deteriorates due to the model uncertainty which encompasses the modeling errors and the joint friction. Moreover, the estimation error is significant when MOB is applied to high-dimensional floating-base humanoids, which prevents the estimated external joint torque from being used for force control or collision detection in the real humanoid robot. In this paper, the pure external joint torque estimation method named MOB-Net, is proposed for humanoids. MOB-Net learns the model uncertainty torque and calibrates the estimated signal of MOB, substantially reducing the estimation errors of MOB. The external joint torque can be estimated in the generalized coordinate including whole-body and virtual joints of the floating-base robot with only internal sensors (an IMU on the pelvis and encoders in the joints). Furthermore, MOB-Net shows more robust performance for the unseen data compared to the end-to-end learning method, and the robustness of MOB-Net is validated through extensive simulations, real robot experiments, and ablation studies. Finally, various collision handling scenarios are presented to show the versatility of MOB-Net: contact wrench feedback control for locomotion, collision detection, and collision reaction for safety.

Problem

• Momentum Observer
• Estimation Error
• Robot Force Estimation

Innovation

• MOB-Net
• Force Estimation
• Sensorless Perception