GRIP: A General Robotic Incremental Potential Contact Simulation Dataset for Unified Deformable-Rigid Coupled Grasping

2025-03-06

AI Summary

Existing large-scale robotic grasping datasets largely neglect deformable objects due to the absence of scalable, robust deformation simulation pipelines, hindering progress in soft-grasping and compliant manipulation models. Method: We introduce the first large-scale simulation dataset for rigid–soft coupled grasping, pioneering the large-scale application of the Incremental Potential Contact (IPC) physics engine to robotic grasping data generation—enabling penetration-free, inversion-free, and computationally efficient simulation of interactions between soft/hard grippers and deformable objects. We further propose a fully automated multi-environment grasping generation and evaluation pipeline, covering 1,200 parametric objects and 100,000 grasp poses. Contribution/Results: Our approach achieves up to 48× speedup over baseline methods, enabling downstream tasks such as neural grasp generation and stress-field prediction. It significantly improves model generalization under complex, large-deformation interaction scenarios.

Technology Category

• Intelligent Robots: Manipulation
• Search and Optimization: Sampling/Simulation-based Search
• Computer Vision: Biometrics, Face, Gesture & Pose

Application Category

• Web Mining and Content Analysis: Web data generation and simulation
• User Modeling, Personalization and Recommendation: User modeling and simulation for interactive and conversational systems
• Economics, Online Markets and Human Computation: Data quality aspects of human-annotated datasets

Abstract

Grasping is fundamental to robotic manipulation, and recent advances in large-scale grasping datasets have provided essential training data and evaluation benchmarks, accelerating the development of learning-based methods for robust object grasping. However, most existing datasets exclude deformable bodies due to the lack of scalable, robust simulation pipelines, limiting the development of generalizable models for compliant grippers and soft manipulands. To address these challenges, we present GRIP, a General Robotic Incremental Potential contact simulation dataset for universal grasping. GRIP leverages an optimized Incremental Potential Contact (IPC)-based simulator for multi-environment data generation, achieving up to 48x speedup while ensuring efficient, intersection- and inversion-free simulations for compliant grippers and deformable objects. Our fully automated pipeline generates and evaluates diverse grasp interactions across 1,200 objects and 100,000 grasp poses, incorporating both soft and rigid grippers. The GRIP dataset enables applications such as neural grasp generation and stress field prediction.

Problem

• Lack of datasets for deformable-rigid coupled grasping simulations
• Need for scalable, robust simulation pipelines for compliant grippers
• Development of generalizable models for soft manipulands and grippers

Innovation

• Optimized IPC-based simulator for fast simulations
• Automated pipeline for diverse grasp interactions
• Supports both soft and rigid gripper simulations