Force Generative Imitation Learning: Bridging Position Trajectory and Force Commands through Control Technique

2026-02-06

IEEE Access

AI Summary

This work addresses the challenge in contact-rich tasks where position trajectories are readily available but desired force commands are difficult to specify, and existing foundation models struggle to generalize across different robotic hardware. To overcome this, the authors propose a memoryless force generation model that directly maps given position trajectories to force control signals, augmented with a feedback control mechanism to enhance stability. By integrating force-generation imitation learning, memoryless neural network modeling, and closed-loop feedback, the approach achieves stable force tracking on unseen trajectories in real-world robotic writing tasks. This significantly improves the system’s generalization capability and control robustness across diverse robotic platforms.

Technology Category

• Intelligent Robots: Behavior Learning & Control
• Machine Learning: Imitation Learning & Inverse Reinforcement Learning
• Natural Language Processing: Generation

Application Category

• User Modeling, Personalization and Recommendation: User modeling and simulation for interactive and conversational systems
• Social Networks and Social Media: Generative AI / large language models and their impact on social systems
• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs

Abstract

In contact-rich tasks, while position trajectories are often easy to obtain, appropriate force commands are typically unknown. Although it is conceivable to generate force commands using a pretrained foundation model such as Vision-Language-Action (VLA) models, force control is highly dependent on the specific hardware of the robot, which makes the application of such models challenging. To bridge this gap, we propose a force generative model that estimates force commands from given position trajectories. However, when dealing with unseen position trajectories, the model struggles to generate accurate force commands. To address this, we introduce a feedback control mechanism. Our experiments reveal that feedback control does not converge when the force generative model has memory. We therefore adopt a model without memory, enabling stable feedback control. This approach allows the system to generate force commands effectively, even for unseen position trajectories, improving generalization for real-world robot writing tasks.

Problem

• force control
• generative imitation learning
• position trajectory
• contact-rich tasks
• robot generalization

Innovation

• Force Generative Imitation Learning
• Feedback Control
• Memoryless Model
• Force Command Generation
• Contact-Rich Manipulation