position-based compliance control

This work addresses the lack of reusable, cross-platform compliant control infrastructure in existing robotic software, which hinders unified algorithm implementation and high-level interfacing. The authors propose a robot-agnostic, modular compliance control framework that decouples controller infrastructure from control laws via a plugin architecture. It supports variable impedance control in both joint and Cartesian spaces and, for the first time, enables online adaptation of the primary compliance direction according to task geometry—overcoming the limitations of fixed coordinate frames. Built upon the ROS ecosystem, the framework leverages Pinocchio to parse URDF models for kinematic and dynamic computations and employs runtime plugin loading with generic wrappers to interface heterogeneous hardware. Real-world and simulated experiments demonstrate significant performance improvements in contact-intensive tasks and seamless transferability across multiple robotic arms.

To address poor compliance and low torque control accuracy in high-friction commercial dual-arm robots (e.g., Kinova Gen3) for human–robot collaborative scenarios, this paper proposes a real-time impedance control method. The approach features: (1) a novel smooth interpolation-based compliant control architecture that jointly operates in task and joint spaces; and (2) a model-free friction observer enabling online friction disturbance compensation without precise dynamic modeling. Implemented on ROS2, the system integrates MoveIt! for motion planning and supports high-frequency, closed-loop streaming execution of trajectory commands. Experiments demonstrate sub-centimeter (<1 cm) peg-in-hole accuracy even under strong friction, achieving both high-robustness trajectory tracking and high-fidelity compliant response. The framework enables real-time closed-loop execution of learned, optimized, and teleoperated trajectories.

In human–robot collaboration (HRC), conventional controllers struggle to adapt to unknown physical interactions at the end-effector and robot body, compromising both safety and task continuity. Method: This paper proposes a two-layer compliant control framework integrating modified Cartesian impedance control with a dynamical systems (DS)-based motion generator. It innovatively unifies null-space impedance control with DS-driven Cartesian impedance, enabling autonomous trajectory recovery following online physical interactions (e.g., tool exchange) while ensuring whole-body passive safety. Contribution/Results: Based on port-Hamiltonian modeling and passivity analysis, the approach is experimentally validated on a KUKA LWR IV+ robot: peak joint torque during unexpected contact decreases by 42%, Cartesian tracking error remains below 1.8 mm, and strict passivity is guaranteed—significantly enhancing HRI safety and task robustness.

This study addresses the challenge of balancing precise tracking with compliant interaction in vision-action policies for contact-rich manipulation, proposing the Imp-ACT method. Imp-ACT integrates direction-dependent stiffness modulation into demonstration collection by dynamically adjusting stiffness along motion directions via a self-tuning impedance controller during teleoperation. Building upon the ACT architecture, it fuses visual, proprioceptive, and torque observations for end-to-end learning, enabling a Transformer to jointly predict poses, actions, and adaptive stiffness. Experimental results demonstrate that Imp-ACT eliminates the need for manual stiffness selection or offline trajectory reconstruction. In wiping tasks, contact force oscillations are reduced by approximately 29-fold, while orthogonal forces during peg insertion decrease by 43%. The method effectively constrains interaction forces while maintaining high task success rates.

Visual motor policies commonly neglect adaptive compliance modulation, leading to poor trade-offs between contact force suppression and trajectory tracking accuracy. Method: We propose the Adaptive Compliance Policy (ACP) framework, which learns spatiotemporal compliance configurations directly from human demonstrations—replacing fixed or pre-specified stiffness assumptions with task-driven, online compliance adaptation. ACP integrates diffusion-model-guided policy learning, demonstration-based compliance estimation, and joint visual–tactile representation modeling. Contribution/Results: Evaluated on contact-intensive manipulation tasks, ACP achieves over 50% performance improvement over state-of-the-art methods, significantly enhancing robotic robustness in uncertain, unstructured environments through compliant, force-aware interaction.

本文通过四通道双边遥操作解决了机器人无法从示教中学习顺应性的问题，使机器人能够根据指令调整接触力。

This study addresses the performance limitations of contact-rich manipulation caused by mismatches between preset stiffness or directions in traditional control and environmental constraints. We propose a proprioception-based reflex strategy trained in simulation with a frozen execution layer. By leveraging interaction primitives such as springs and planes alongside state-history mapping, the method translates task commands into joint targets, decoupling contact responses from command generation without requiring direct force or geometric measurements. Experimental results demonstrate that this approach maintains low contact forces during box lifting and outperforms baselines in surface following. Furthermore, it improves peg-in-hole insertion success rates by 36–58% while reducing contact forces by approximately 50%, thereby providing robust low-level execution capabilities for high-level planning.

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.

This study addresses the challenges of force perception and multi-contact load regulation in dexterous manipulation. To this end, it proposes the HACo strategy, which integrates fingertip tactile sensing with joint torque measurements to achieve closed-loop active compliance control. Specifically, a compliant grounding module is designed to facilitate supervised learning of force-regulation actions without requiring explicit online contact modeling. Furthermore, a gated tactile cross-attention mechanism is introduced to efficiently align cross-modal features, complemented by a compliant teleoperation pipeline for data generation. Real-world benchmark evaluations demonstrate that the proposed approach achieves an average success rate of 83%, significantly outperforming existing baseline methods.

This study addresses the challenge of balancing external force responsiveness with motion accuracy and the lack of directional compliance control during physical interactions in humanoid robots. To this end, we propose a hierarchical reinforcement learning framework that integrates end-effector directionally adjustable compliance with a root-selective compliance mechanism. A high-level policy modulates a low-level whole-body tracking controller, leveraging historical proprioceptive information to achieve online contact force estimation and dynamic adjustment of compliance parameters. Both simulation and real-world experiments validate the effectiveness of the proposed framework across complex mobile manipulation tasks, including directional stiffness control, disturbance-rejection tracking, and collaborative object transportation.