Method and system for preserving privacy for cloud-based manufacturing analysis services

Patent metadata

Field	Value
Publication number	US-10933530-B2
Application number	US-201716325414-A
Country	US
Kind code	B2
Filing date	Aug 16, 2017
Priority date	Aug 17, 2016
Publication date	Mar 2, 2021
Grant date	Mar 2, 2021

Abstract

A system for analyzing geometric properties for an object includes designing the object in a first computer process and producing information relating to the geometric properties of the object, and receiving the information in a second computer processor which identifies a first portion of the geometric property information as masked or private and second portion identified as public or shared, analysis is performed by the second processor on the public/shared portion of the geometric property information. An output based on the analysis may be provided to an industrial system performing processes on the object. A binary privacy label may be assigned to each triangle in a set of triangles representing the surfaces of the object in a 3D object mesh. The privacy label denotes an associated triangle as being private or shared. The system may be used to produce a set of planned grasps for a robotic gripper.

First claim

What is claimed is:

1. A system for analyzing geometric properties of an object for robotic grasp planning, comprising:
a first computer processor configured to design the object and produce information relating to the geometric properties of the object;
a second computer processor in communication with the first computer processor, the second processor configured to receive the information relating to the geometric properties of the object;
wherein a first portion of the information relating to the geometric properties represents a private/masked portion of the object that conceals at least one aspect of the geometric properties of the object, and a second portion of the information relating to the geometric properties of the object represents a sharable/unmasked portion of the object, and the second processor is configured to analyze the unmasked portion of the object and provide output information of the analysis to an industrial system;
a privacy labeling tool configured to run on the first processor, the labeling tool adapted to:
display a 3D mesh representation of the object, the mesh comprising a plurality of vertices and a plurality of triangles defined by the connection of three of the vertices;
receive one or more commands from a user defining hounding boxes that contain at least one of the triangles of the mesh that are identified as masked portions of the object; and
label each triangle of the mesh with a binary privacy label to indicate if the triangle should be private/masked or sharable/unmasked; and

a candidate grasp generation tool configured to run on the second processor, the candidate grasp generation tool configured to identify a set of candidate grasps for a robot to grasp the object, each grasp based on two contact points on a surface of the object mesh, wherein each of the two contact points are associated with a center of a corresponding triangle in the sharable/unmasked portion of the object mesh.

2. The system of claim 1 , further comprising:
a privacy masking tool configured to run on the first processor, the privacy masking tool configured to apply a masking function, the masking function configured to mask or hide the triangles marked as private based on the binary privacy label of each triangle in the mesh, wherein a triangle having a binary privacy value of 1 is private and a triangle having a binary privacy value of 0 is sharable.

3. The system of claim 2 , wherein the first processor computes stable poses of the object and transmits those to the second processor.

4. The system of claim 2 , wherein the masking function is configured to create a bounding box around regions of the mesh that contain triangles associated with masked portions of the object mesh.

5. The system of claim 3 , wherein the masking function is configured to replace connected components associated with masked portions of the object mesh with a convex hull of each connected component.

6. The system of claim 1 , wherein the candidate grasp generation tool is further configured to:
rank the candidate grasps in order of robustness to perturbations.

7. The system of claim 6 , wherein the candidate grasp generation tool is further configured to subdivide the at least one triangle in the unmasked portion of the object mesh by primal triangular quadrisection.

8. The system of claim 6 , wherein the candidate grasp generation tool is further configured to:
determine for candidate grasps for each stable pose of the object:
whether operation of the robotic gripper associated with each candidate grasp results in a collision of the gripper with a portion of the object; and
if there is no collision for a particular candidate grasp, determine a robustness measure of the candidate grasp indicative of whether operation of the gripper to contact the object at the associated contact points for the associated stable pose of the object will result in successful grasp of the object.

9. The system of claim 8 , wherein the robustness measure is based on a probability of force closure under object pose, robotic gripper pose, and friction uncertainty.

10. The system of claim 6 , wherein the candidate grasp generation tool is further configured to:
generate information indicative of an improvement of candidate grasp quality, based on a suggested reduction in the size of the private/masked portion of the object.

11. The system of claim 1 , wherein the second computer processor is configured to provide output information relating to fixturing of a part in the industrial system.

12. The system of claim 1 , wherein the second processor is associated with a grasp planning as a service (GPaaS) online application.

13. A method of analyzing geometric properties of an object for robotic grasp planning, comprising:
in a computer processor, receiving design information relating to a first portion of the received designed information as a private/masked portion of the object and to a second portion of the received design information as a shared/unmasked portion of the object, wherein the private/masked portion of the object conceals at least one aspect of the geometric properties of the object;
in the processor, analyzing the first and second portions of the design information relating to the geometric properties of the object; and
in the processor, producing an output representative of a property corresponding to the geometric properties of the object, based on the unmasked portion of the object;
providing the output representative of the property to an industrial system;
in the processor, receiving a command from a user, the command operative to identify at least one proprietary zone comprising at least one triangle of the object mesh;
creating a binary privacy label for each triangle in the object mesh, wherein the binary privacy label has a first value representative of a private/masked triangle and a second value representative of a sharable/unmasked triangle; and
in the processor, identifying a set of grasps based on three-dimensional mesh for a robotic gripper and sharable/unmasked triangles of the object mesh.

14. The method of claim 13 , further comprising:
in the computer processor, representing the geometric properties of the object as a three-dimensional mesh comprising a plurality of vertices and a plurality of triangles representative of surfaces of the object, each triangle defined by three of the plurality of vertices.

15. The method of claim 14 , further comprising:
in the computer processor, determining a set of stable poses for the mesh relative an infinite planar work surface.

16. The method of claim 14 , further comprising:
in the processor, assigning a quality metric to each grasp in the set of grasps; and
outputting a set of planned grasps, wherein each planned grasp has a quality metric that exceeds a predetermined threshold.

Assignees

• Siemens Ag
• Univ California

Inventors

• Goldberg Kenneth Yigael
• Chandra Ramu
• Pokorny Florian Till
• Mahler Jeffrey Brian
• Aparicio Ojea Juan L

Classifications

• G05B2219/40113

  Task planning · CPC title

• B25J9/1602

  characterised by the control system, structure, architecture · CPC title

• G06T17/10

  Constructive solid geometry [CSG] using solid primitives, e.g. cylinders, cubes · CPC title

• G05B2219/39543

  Recognize object and plan hand shapes in grasping movements · CPC title

• G06T17/20

  Finite element generation, e.g. wire-frame surface description, {tesselation} · CPC title