Research Analyzer
← Back ICRA 2026

A Passivity-Based Framework for Dynamic Arbitration between Trajectory and Force Tracking using Human Demonstration

Yeoil Yun, Youngwuk Kim, Junchul Gwak, Hyungpil Moon, Hyouk Ryeol Choi, Ja Choon Koo

PDF

AI summary

Key figure (auto-extracted from paper)
A one-shot learning-from-demonstration framework uses the operator’s grip force to dynamically blend trajectory and force tracking controllers while guaranteeing energetic stability through a dual-layer passivity mechanism.
Learning from demonstration Passivity-based control Grip force arbitration Impedance/admittance control Safe human-robot interaction Redundant manipulators

Problem

Learning from demonstration for contact-rich tasks struggles to balance precise trajectory tracking with accurate force reproduction, often failing when environmental geometry deviates or when switching control modes causes instability.

Approach

The method converts the operator’s physical grip force into a continuous blending weight that smoothly transitions between impedance and admittance controllers, while a dual-layer passivity system safely manages energy injection to prevent instability.

Key results

  • Grip force used as a continuous, intuitive signal for dynamic control arbitration
  • Robust replication of demonstrated interaction forces under positional uncertainty
  • Dual-layer passivity assurance with dynamic energy arbitration between tank and null-space dissipation
  • Experimental validation on a 7-DOF manipulator showing superior safety and performance over conventional impedance control

Why it matters

Provides a practical, sensor-light method for safely transferring complex contact-rich manipulation skills to robots without intrusive physiological measurements or precise environmental modeling.

Abstract

Learning from Demonstration (LfD) for contact- rich tasks faces a fundamental challenge: arbitrating between tracking a demonstrated trajectory and reproducing an inter- action force. This paper introduces a novel one-shot LfD frame- work that resolves this conflict by leveraging the operator’s grip force as an intuitive, continuous signal for arbitration. This signal allows the controller to seamlessly transition between a trajectory-tracking impedance controller and a force-tracking admittance controller, prioritizing path accuracy when the demonstrated grip was light and interaction force fidelity when it was firm. To ensure verifiably safe interaction, the adaptive control law is integrated within a dual-layer passivity assurance framework. This mechanism intelligently distributes potentially non-passive energy between an energy tank and adaptive null-space dissipation to guarantee energetic stability. The proposed framework was experimentally validated on a 7-DOF manipulator, demonstrating that the controller autonomously reproduces interaction forces and shows significant robustness against environmental position uncertainties, a scenario where conventional impedance controllers can fail.

Index terms

Physical Human-Robot Interaction Learning from Demonstration Intention Recognition

Related papers