Abstract:Robots are increasingly being deployed not only in workplaces but also in households. Effectively execute of manipulation tasks by robots relies on variable impedance control with contact forces. Furthermore, robots should possess adaptive capabilities to handle the considerable variations exhibited by different robotic tasks in dynamic environments, which can be obtained through human demonstrations. This paper presents a learning-from-demonstration framework that integrates force sensing and motion information to facilitate variable impedance control. The proposed approach involves the estimation of full stiffness matrices from human demonstrations, which are then combined with sensed forces and motion information to create a model using the non-parametric method. This model allows the robot to replicate the demonstrated task while also responding appropriately to new task conditions through the use of the state-dependent stiffness profile. Additionally, a novel tank based variable impedance control approach is proposed to ensure passivity by using the learned stiffness. The proposed approach was evaluated using two virtual variable stiffness systems. The first evaluation demonstrates that the stiffness estimated approach exhibits superior robustness compared to traditional methods when tested on manual datasets, and the second evaluation illustrates that the novel tank based approach is more easily implementable compared to traditional variable impedance control approaches.

What problem does this paper attempt to address?

The paper attempts to address the problem of how to learn variable impedance skills through human demonstration while ensuring the passivity of the system (i.e., the system does not generate additional energy, thereby ensuring the safety and stability of the interaction) when a robot performs tasks. Specifically, the paper focuses on: 1. **Variable Impedance Control through Demonstration Learning**: The robot needs to adaptively adjust its impedance (stiffness and damping) in different dynamic environments to complete complex operational tasks. Traditional fixed coding methods are difficult to achieve this in unknown and unstructured environments. Therefore, the paper proposes a method to learn variable impedance skills from human demonstrations. 2. **Ensuring System Passivity**: In physical human-robot interaction, system passivity is very important because it can prevent the robot from generating additional energy during the interaction, thereby ensuring the safety and stability of the interaction. The paper designs a new Lyapunov function to ensure that the system remains passive when using the learned impedance parameters. ### Main Contributions of the Paper: 1. **Effective Stiffness Estimation Method**: A method is proposed to estimate the stiffness matrix from demonstrations, regardless of whether the damping information is known. 2. **New Lyapunov Function**: A new Lyapunov function is designed, and an easily implementable stability condition is derived to ensure system passivity when using the learned stiffness parameters. 3. **Experimental Validation**: Through virtual simulations and experiments with actual robots, the robustness and practicality of the proposed framework in learning stiffness from demonstrations and applying it to variable impedance control are demonstrated. ### Structure of the Paper: - **Introduction**: Introduces the research background and motivation. - **Related Work**: Reviews existing research, especially methods for learning variable impedance skills from human demonstrations. - **Proposed Method**: Details the method for estimating stiffness from demonstrations and how to apply these stiffness parameters to variable impedance control, ensuring system passivity. - **Experimental Results**: Shows the results of simulations and actual robot experiments, validating the effectiveness of the method. - **Conclusion**: Summarizes the main findings and contributions of the paper. Through these contributions, the paper provides an effective and safe method for robots to perform complex tasks in dynamic environments.

Learning Variable Impedance Skills from Demonstrations with Passivity Guarantee

Learning and Generalizing Variable Impedance Manipulation Skills from Human Demonstrations

A neural network based framework for variable impedance skills learning from demonstrations

Variable Stiffness Control of a Soft Rehabilitation Robot with Sliding Mode Method

Variable Impedance Learning from Human Demonstrations

Robot Variable Impedance Control and Generalizing from Human–Robot Interaction Demonstrations

Effective Learning and Online Modulation for Robotic Variable Impedance Skills.

Learning Compliant Manipulation Tasks from Force Demonstrations

A Passivity-Based Variable Impedance Controller for Incremental Learning of Periodic Interactive Tasks

Efficient Force Control Learning System For Industrial Robots Based On Variable Impedance Control

Human-Robot Variable Impedance Skills Transfer Learning Based on Dynamic Movement Primitives

Robot Variable Impedance Skill Transfer and Learning Framework Based on a Simplified Human Arm Impedance Model

Efficient Learning Variable Impedance Control for Industrial Robots

Learning compliant dynamical system from human demonstrations for stable force control in unknown environments

Learning Variable Impedance Control via Inverse Reinforcement Learning for Force-Related Tasks

Robot Learning from Human Demonstration of Peg-In-Hole Task

Implementation of Impedance Control for Lower-Limb Rehabilitation Robots

A DMP-based Framework for Efficiently Generating Complete Stiffness Profiles of Human-like Variable Impedance Skill from Demonstrations

Development of an Impedance-based Human-robot Hybrid Interaction System with RNN Force Estimator on Visual Tasks

A Framework for Autonomous Impedance Regulation of Robots Based on Imitation Learning and Optimal Control

Simultaneously Encoding Movement and sEMG-Based Stiffness for Robotic Skill Learning