PinchBot: Long-Horizon Deformable Manipulation with Guided Diffusion Policy

2025-07-23

AI Summary

This paper addresses the long-horizon, multimodal robotic manipulation of deformable objects—specifically clay—by proposing a novel, pinch-only autonomous pottery shaping method. To tackle the challenges of high-dimensional state representation and long-range deformation control, we introduce a target-conditioned diffusion policy framework. It integrates a pre-trained 3D point cloud encoder for deformation state encoding, and incorporates task-progress prediction and collision-aware action projection to enable precise, long-horizon deformation path planning and execution. The method is validated in both simulation and on a real robotic platform, successfully generating diverse simple ceramic forms with demonstrated effectiveness and robustness. To foster reproducibility and community advancement, we publicly release a demonstration dataset and experimental videos, establishing a new paradigm and benchmark resource for deformable object manipulation research.

Technology Category

• Intelligent Robots: Manipulation
• Computer Vision: Diffusion Models for Vision
• Planning, Routing, and Scheduling: Replanning and Plan Repair

Application Category

• Responsible Web: Machine-in-the-loop, human agency and autonomy
• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications
• Graph Algorithms and Modeling for the Web: Efficient manipulation of static and dynamic Web-related graphs

Abstract

Pottery creation is a complicated art form that requires dexterous, precise and delicate actions to slowly morph a block of clay to a meaningful, and often useful 3D goal shape. In this work, we aim to create a robotic system that can create simple pottery goals with only pinch-based actions. This pinch pottery task allows us to explore the challenges of a highly multi-modal and long-horizon deformable manipulation task. To this end, we present PinchBot, a goal-conditioned diffusion policy model that when combined with pre-trained 3D point cloud embeddings, task progress prediction and collision-constrained action projection, is able to successfully create a variety of simple pottery goals. For experimental videos and access to the demonstration dataset, please visit our project website: https://sites.google.com/andrew.cmu.edu/pinchbot/home.

Problem

• Develop robotic system for pinch-based pottery creation
• Address multi-modal long-horizon deformable manipulation challenges
• Achieve diverse pottery goals using diffusion policy model

Innovation

• Goal-conditioned diffusion policy model
• Pre-trained 3D point cloud embeddings
• Collision-constrained action projection