# Dex-Net 3.0: Computing Robust Robot Vacuum Suction Grasp Targets in Point Clouds using a New Analytic Model and Deep Learning

arXiv: https://arxiv.org/abs/1709.06670 · HTML: https://mlsanity.com/paper/1709.06670/

## Key facts

- Cited by 15 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 19 Sep 2017.

## Built on by (most cited first, 15 of 15)

- [FlowBot3D: Learning 3D Articulation Flow to Manipulate Articulated Objects](https://mlsanity.com/paper/2205.04382/) (2022)
- [Sample-Efficient Safety Assurances using Conformal Prediction](https://mlsanity.com/paper/2109.14082/) (2021)
- [MetaGraspNet: A Large-Scale Benchmark Dataset for Scene-Aware Ambidextrous Bin Picking via Physics-based Metaverse Synthesis](https://mlsanity.com/paper/2208.03963/) (2022)
- [Deep Reinforcement Learning for Robotic Pushing and Picking in Cluttered Environment](https://mlsanity.com/paper/2302.10717/) (2023)
- [GKNet: grasp keypoint network for grasp candidates detection](https://mlsanity.com/paper/2106.08497/) (2021)
- [Fast GraspNeXt: A Fast Self-Attention Neural Network Architecture for Multi-task Learning in Computer Vision Tasks for Robotic Grasping on the Edge](https://mlsanity.com/paper/2304.11196/) (2023)
- [Transferring Experience from Simulation to the Real World for Precise Pick-And-Place Tasks in Highly Cluttered Scenes](https://mlsanity.com/paper/2101.04781/) (2021)
- [Learning to Efficiently Plan Robust Frictional Multi-Object Grasps](https://mlsanity.com/paper/2210.07420/) (2022)
- [Model-free Grasping with Multi-Suction Cup Grippers for Robotic Bin Picking](https://mlsanity.com/paper/2307.16488/) (2023)
- [Kit-Net: Self-Supervised Learning to Kit Novel 3D Objects into Novel 3D Cavities](https://mlsanity.com/paper/2107.05789/) (2021)
- [VGQ-CNN: Moving Beyond Fixed Cameras and Top-Grasps for Grasp Quality Prediction](https://mlsanity.com/paper/2203.04874/) (2022)
- [Reinforcement Learning for Picking Cluttered General Objects with Dense Object Descriptors](https://mlsanity.com/paper/2304.10108/) (2023)
- [Dynamic Knowledge Graphs as Semantic Memory Model for Industrial Robots](https://mlsanity.com/paper/2101.01099/) (2021)
- [Precise Object Placement with Pose Distance Estimations for Different Objects and Grippers](https://mlsanity.com/paper/2110.00992/) (2021)
- [A Review of Scene Representations for Robot Manipulators](https://mlsanity.com/paper/2301.11275/) (2022)

## Abstract (arXiv)

Vacuum-based end effectors are widely used in industry and are often preferred over parallel-jaw and multifinger grippers due to their ability to lift objects with a single point of contact. Suction grasp planners often target planar surfaces on point clouds near the estimated centroid of an object. In this paper, we propose a compliant suction contact model that computes the quality of the seal between the suction cup and local target surface and a measure of the ability of the suction grasp to resist an external gravity wrench. To characterize grasps, we estimate robustness to perturbations in end-effector and object pose, material properties, and external wrenches. We analyze grasps across 1,500 3D object models to generate Dex-Net 3.0, a dataset of 2.8 million point clouds, suction grasps, and grasp robustness labels. We use Dex-Net 3.0 to train a Grasp Quality Convolutional Neural Network (GQ-CNN) to classify robust suction targets in point clouds containing a single object. We evaluate the resulting system in 350 physical trials on an ABB YuMi fitted with a pneumatic suction gripper. When evaluated on novel objects that we categorize as Basic (prismatic or cylindrical), Typical (more complex geometry), and Adversarial (with few available suction-grasp points) Dex-Net 3.0 achieves success rates of 98%, 82%, and 58% respectively, improving to 81% in the latter case when the training set includes only adversarial objects. Code, datasets, and supplemental material can be found at http://berkeleyautomation.github.io/dex-net .