Processing

Please wait...

Settings

Settings

Goto Application

1. WO2020194187 - HYBRID MACHINE LEARNING-BASED SYSTEMS AND METHODS FOR TRAINING AN OBJECT PICKING ROBOT WITH REAL AND SIMULATED PERFORMANCE DATA

Publication Number WO/2020/194187
Publication Date 01.10.2020
International Application No. PCT/IB2020/052757
International Filing Date 24.03.2020
IPC
B25J 9/16 2006.01
BPERFORMING OPERATIONS; TRANSPORTING
25HAND TOOLS; PORTABLE POWER-DRIVEN TOOLS; HANDLES FOR HAND IMPLEMENTS; WORKSHOP EQUIPMENT; MANIPULATORS
JMANIPULATORS; CHAMBERS PROVIDED WITH MANIPULATION DEVICES
9Programme-controlled manipulators
16Programme controls
G05B 13/04 2006.01
GPHYSICS
05CONTROLLING; REGULATING
BCONTROL OR REGULATING SYSTEMS IN GENERAL; FUNCTIONAL ELEMENTS OF SUCH SYSTEMS; MONITORING OR TESTING ARRANGEMENTS FOR SUCH SYSTEMS OR ELEMENTS
13Adaptive control systems, i.e. systems automatically adjusting themselves to have a performance which is optimum according to some preassigned criterion
02electric
04involving the use of models or simulators
CPC
B25J 9/1612
BPERFORMING OPERATIONS; TRANSPORTING
25HAND TOOLS; PORTABLE POWER-DRIVEN TOOLS; MANIPULATORS
JMANIPULATORS; CHAMBERS PROVIDED WITH MANIPULATION DEVICES
9Programme-controlled manipulators
16Programme controls
1612characterised by the hand, wrist, grip control
B25J 9/163
BPERFORMING OPERATIONS; TRANSPORTING
25HAND TOOLS; PORTABLE POWER-DRIVEN TOOLS; MANIPULATORS
JMANIPULATORS; CHAMBERS PROVIDED WITH MANIPULATION DEVICES
9Programme-controlled manipulators
16Programme controls
1628characterised by the control loop
163learning, adaptive, model based, rule based expert control
B25J 9/1671
BPERFORMING OPERATIONS; TRANSPORTING
25HAND TOOLS; PORTABLE POWER-DRIVEN TOOLS; MANIPULATORS
JMANIPULATORS; CHAMBERS PROVIDED WITH MANIPULATION DEVICES
9Programme-controlled manipulators
16Programme controls
1656characterised by programming, planning systems for manipulators
1671characterised by simulation, either to verify existing program or to create and verify new program, CAD/CAM oriented, graphic oriented programming systems
G05B 2219/39473
GPHYSICS
05CONTROLLING; REGULATING
BCONTROL OR REGULATING SYSTEMS IN GENERAL; FUNCTIONAL ELEMENTS OF SUCH SYSTEMS; MONITORING OR TESTING ARRANGEMENTS FOR SUCH SYSTEMS OR ELEMENTS
2219Program-control systems
30Nc systems
39Robotics, robotics to robotics hand
39473Autonomous grasping, find, approach, grasp object, sensory motor coordination
G06F 17/18
GPHYSICS
06COMPUTING; CALCULATING; COUNTING
FELECTRIC DIGITAL DATA PROCESSING
17Digital computing or data processing equipment or methods, specially adapted for specific functions
10Complex mathematical operations
18for evaluating statistical data ; , e.g. average values, frequency distributions, probability functions, regression analysis
G06N 20/00
GPHYSICS
06COMPUTING; CALCULATING; COUNTING
NCOMPUTER SYSTEMS BASED ON SPECIFIC COMPUTATIONAL MODELS
20Machine learning
Applicants
  • ABB SCHWEIZ AG [CH]/[CH]
Inventors
  • HUANG, Jinmiao
  • MARTINEZ, Carlos
  • CHOI, Sangeun
  • FUHLBRIGGE, Thomas A.
Priority Data
16/366,70027.03.2019US
Publication Language English (EN)
Filing Language English (EN)
Designated States
Title
(EN) HYBRID MACHINE LEARNING-BASED SYSTEMS AND METHODS FOR TRAINING AN OBJECT PICKING ROBOT WITH REAL AND SIMULATED PERFORMANCE DATA
(FR) SYSTÈMES BASÉS SUR L'APPRENTISSAGE AUTOMATIQUE HYBRIDE ET PROCÉDÉS D'ENTRAÎNEMENT DE ROBOT DE PRISE D'OBJETS AVEC DONNÉES DE PERFORMANCE RÉELLES ET SIMULÉES
Abstract
(EN)
For training an object picking robot with real and simulated grasp performance data, grasp locations on an object are assigned based on object physical properties. A simulation experiment for robot grasping is performed using a first set of assigned locations. Based on simulation data from the simulation, a simulated object grasp quality of the robot is evaluated for each of the assigned locations. A first set of candidate grasp locations on the object is determined based on data representative of simulated grasp quality from the evaluation. Based on sensor data from an actual experiment for the robot grasping using each of the candidate grasp locations, an actual object grasp quality is evaluated for each of the candidate locations.
(FR)
Selon l'invention, pour entraîner un robot de prise d'objets avec des données de performance de saisie réelles et simulées, des emplacements de saisie sur un objet sont attribués en fonction de propriétés physiques d'objet. Une expérience de simulation pour la saisie par le robot est effectuée en utilisant un premier ensemble d'emplacements attribués. En fonction de données de simulation provenant de la simulation, une qualité de saisie d'objet simulée du robot est évaluée pour chacun des emplacements attribués. Un premier ensemble d'emplacements de saisie candidats sur l'objet est déterminé en fonction de données représentatives de la qualité de saisie simulée à partir de l'évaluation. En fonction de données de capteur provenant d'une expérience réelle pour la saisie par le robot en utilisant chacun des emplacements de saisie candidats, une qualité de saisie d'objet réel est évaluée pour chacun des emplacements candidats.
Also published as
Latest bibliographic data on file with the International Bureau