IMPACT: Learning Internal-Model Predictive Control for Forceful Robotic Manipulation

2026-06-09

AI Summary

This work addresses the limited generalization of existing learning-based robotic manipulation methods in high-force interaction tasks and their frequent reliance on specialized force or tactile sensors, which increases system complexity. The authors propose IMPACT, a novel framework that, for the first time, integrates internal models with model predictive control by decoupling the task into two stages: high-level planning and internal-model-based predictive control. By implicitly learning to predict interaction forces, IMPACT achieves precise force regulation without requiring external force sensors. The approach synergistically combines internal model learning, end-to-end policy training, and impedance control, significantly improving task success rates in both simulation and real-world experiments. It demonstrates strong generalization to unseen object masses while simultaneously enhancing operational safety and energy efficiency.

Abstract

Real-world robotic manipulation tasks often involve forceful interactions with the environment, such as using tools of varying weights, transporting objects with different masses, and performing contact-rich tasks like table wiping. Previous learning-based approaches typically employ imitation learning policies that output target end-effector poses tracked by low-level impedance controllers. In these systems, forceful interactions are either implicitly realized through steady-state tracking errors or explicitly commanded using wrist force/torque or tactile sensors. However, implicit approaches generalize poorly across object weights, while explicit approaches require specialized hardware and increase system complexity. In this work, we propose IMPACT, a framework that decouples these forceful tasks into task-planning and internal-model-based predictive control. Extensive simulation and real-world experiments demonstrate that the proposed framework achieves higher success rates and improved generalization to unseen object weights, as well as better safety and energy efficiency.

Problem

• forceful manipulation
• generalization
• robotic control
• object weight variation
• contact-rich tasks

Innovation

• internal-model predictive control
• forceful manipulation
• generalization
• robotic control
• model-based planning