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

TL;DR

This paper introduces a force generative model combined with feedback control to accurately produce force commands from position trajectories in contact-rich robotic tasks, enhancing generalization to unseen trajectories.

Contribution

It proposes a novel force generative model integrated with a feedback control mechanism to improve force command generation for unseen trajectories in contact-rich tasks.

Findings

Feedback control fails with memory-based models.

Memoryless models enable stable feedback control.

Enhanced force command generation for unseen trajectories.

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.