CATCH-FORM-ACTer: Compliance-Aware Tactile Control and Hybrid Deformation Regulation-Based Action Transformer for Viscoelastic Object Manipulation

2025-04-11

AI Summary

Tactile manipulation of viscoelastic objects faces core challenges including dynamic parameter mismatch, contact-induced oscillations, and spatiotemporal coupling between force and deformation. Method: This paper proposes a unified framework integrating compliant感知 tactile control, PDE-driven deformation stability regulation, and an Action Chunking Transformer (ACT) architecture. We pioneer the extension of ACT into a real-time, online-tunable, human-like multi-stage force-deformation coordinator—adaptively adjusting stiffness, damping, and diffusion parameters without task- or object-specific manual calibration. The framework synergizes admittance control, learning-from-demonstration (LfD), and multimodal tactile-visual perception. Results: Evaluated on three representative tasks using single- and dual-arm robots, it achieves state-of-the-art force-field tracking performance, improves success rates by 10–20%, and attains sub-millimeter surface deformation accuracy—enabling safe, precise manipulation in industrial, medical, and domestic settings.

Technology Category

• Intelligent Robots: Manipulation
• Humans and AI: Interaction Techniques and Devices
• Computer Vision: Motion & Tracking

Application Category

• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications
• Economics, Online Markets and Human Computation: LLM based quality controls for crowd work
• Responsible Web: Machine-in-the-loop, human agency and autonomy

Abstract

Automating contact-rich manipulation of viscoelastic objects with rigid robots faces challenges including dynamic parameter mismatches, unstable contact oscillations, and spatiotemporal force-deformation coupling. In our prior work, a Compliance-Aware Tactile Control and Hybrid Deformation Regulation (CATCH-FORM-3D) strategy fulfills robust and effective manipulations of 3D viscoelastic objects, which combines a contact force-driven admittance outer loop and a PDE-stabilized inner loop, achieving sub-millimeter surface deformation accuracy. However, this strategy requires fine-tuning of object-specific parameters and task-specific calibrations, to bridge this gap, a CATCH-FORM-ACTer is proposed, by enhancing CATCH-FORM-3D with a framework of Action Chunking with Transformer (ACT). An intuitive teleoperation system performs Learning from Demonstration (LfD) to build up a long-horizon sensing, decision-making and execution sequences. Unlike conventional ACT methods focused solely on trajectory planning, our approach dynamically adjusts stiffness, damping, and diffusion parameters in real time during multi-phase manipulations, effectively imitating human-like force-deformation modulation. Experiments on single arm/bimanual robots in three tasks show better force fields patterns and thus 10%-20% higher success rates versus conventional methods, enabling precise, safe interactions for industrial, medical or household scenarios.

Problem

• Automating manipulation of viscoelastic objects with rigid robots
• Addressing dynamic parameter mismatches and unstable contact oscillations
• Reducing need for object-specific fine-tuning and task calibrations

Innovation

• Combines force-driven admittance and PDE-stabilized loops
• Uses Action Chunking with Transformer for dynamic adjustments
• Learns from human demonstrations for precise force-deformation modulation