VIDP: Variable Impedance Diffusion Policy for Compliant Robot Manipulation from Diverse Demonstrations

2026-08-06

AI Summary

This work addresses the limitations of static compliance in contact-rich manipulation tasks, where dynamic changes in contact constraints are difficult to accommodate and implicit variable impedance strategies cannot be reliably inferred from force-free motion trajectories. To overcome these challenges, the paper proposes a sensorless variable impedance control framework based on imitation learning, which uniquely integrates a Task-Parameterized Direction-Aware Mixture Model (TP-DAMM) with diffusion policies. This approach extracts physically consistent trajectory distributions from diverse demonstrations and decouples geometric adaptation from intent-aware compliance, enabling joint prediction of pose actions and stiffness configurations. Real-world robotic experiments demonstrate that the proposed method significantly improves task success rates, reduces interaction forces compared to high-stiffness controllers, and achieves lower tracking errors than low-stiffness baselines.

Abstract

Contact-rich manipulation requires precise tracking and mechanical compliance, where variable impedance control can improve robustness in task success, whereas static compliance cannot adapt to varying contact constraints. Variable impedance skills can be learned from demonstrations, avoiding complex modeling, but compliance is a hidden variable in force-agnostic kinematic data. While existing methods infer compliance from trajectory variations, these variations may reflect geometric adaptation and not intentional compliance when subject to changing spatial layouts. Therefore, this letter introduces Variable Impedance Diffusion Policy (VIDP), an imitation learning-based variable impedance control framework leveraging a Task-Parameterized Directionality-Aware Mixture Model (TP-DAMM) to extract physically consistent trajectory distributions from diverse demonstrations. By mapping distributions to stiffness profiles, VIDP jointly predicts pose actions and task compliance without force sensors. Real-world experiments show that VIDP significantly outperforms fixed-impedance baselines in task success rate while reducing interaction forces with respect to high stiffness controllers and tracking errors with respect to low stiffness baselines.

Problem

• variable impedance control
• compliant manipulation
• imitation learning
• contact-rich manipulation
• trajectory variation

Innovation

• Variable Impedance Control
• Imitation Learning
• Diffusion Policy
• Task-Parameterized Model
• Compliant Manipulation