US 12,403,613 B2
Process of interacting with objects
Antonio Bicchi, Massa (IT); Gianluca Lentini, Cormano (IT); Manuel Giuseppe Catalano, Pisa (IT); and Giorgio Grioli, Pisa (IT)
Assigned to UNIVERSITA DI PISA, Pisa (IT); and FONDAZIONE ISTITUTO ITALIANO DI TECNOLOGIA, Genoa (IT)
Appl. No. 18/546,286
Filed by UNIVERSITÀ DI PISA, Pisa (IT); and FONDAZIONE ISTITUTO ITALIANO DI TECNOLOGIA, Genoa (IT)
PCT Filed Feb. 7, 2022, PCT No. PCT/IB2022/051027
§ 371(c)(1), (2) Date Aug. 14, 2023,
PCT Pub. No. WO2022/175777, PCT Pub. Date Aug. 25, 2022.
Claims priority of application No. 102021000003821 (IT), filed on Feb. 19, 2021.
Prior Publication US 2024/0116192 A1, Apr. 11, 2024
Int. Cl. B25J 9/16 (2006.01); B25J 13/08 (2006.01); B25J 15/12 (2006.01)
CPC B25J 13/087 (2013.01) [B25J 9/163 (2013.01); B25J 9/1694 (2013.01); B25J 15/12 (2013.01)] 10 Claims
OG exemplary drawing
 
1. A process of interacting with objects comprising
at least one robot comprising:
an end effector of interaction with at least one of said objects;
a driver of said end effector;
sensors for acquiring at least one parameter; said at least one parameter comprising at least one of an environmental parameter and a parameter of interaction of said robot with said object;
a computer for controlling said robot in data connection with said sensor;
an instruction block configured to instruct said robot to move said end effector;
a plurality of learning phases; in each of said learning phases
said instruction block defines a virtual action in which said robot performs a movement of said end effector,
said computer commands said robot to perform a real action according to said virtual action, and
said sensor acquires said at least one parameter during the execution of said real action;
an analysis phase in which said computer
subdivides each of said real actions into an initial scene, a final scene, and
associates with each of said scenes at least one value corresponding to said at least one parameter acquired in said scene;
a reprocessing phase in which said computer creates an action database associating said real actions having substantially the same initial value so as to define a compound action comprising a single initial scene and a plurality of final scenes; said action database associates at least one of said values with each of said scenes;
an execution phase in which
said sensor defines at least one initial factor by detecting said at least one parameter said computer selects in said action database said compound action having said initial value substantially equal to said initial factor and commands said robot to perform an operation substantially equal to a scene subsequent to said initial scene of said compound action; and
said computer commands said robot the next operation to be performed between said scenes associated with said compound action in said action database chosen by comparing
said at least one value associated in said action database with the at least one scene prior to said next scene
with said at least one factor acquired by detecting said at least one parameter in said operation prior to said subsequent operation,
and comprising:
a sub-phase of acquisition in which said sensor defines said initial factor by detecting said at least one parameter,
a identification sub-phase in which said computer selects in said database actions having said initial value substantially equal to said initial factor and commands to said robot the execution of at least one scene subsequent to said initial scene;
at least one development sub-phase in which said robot performs an operation according to said initial scene of said compound action selected in said identification sub-phase and defines a factor by acquiring at least one parameter,
at least one selection sub-phase wherein said computer identifies the next scene to be executed among said scenes of said compound action and commands said robot to perform an operation according to said next scene; and wherein said computer identifies said next scene by comparing said at least one value associated with said at least one scene prior to said next scene with said factor acquired in said at least one prior operation;
wherein said sensor is configured to acquire said movements of said robot by defining instructions necessary for defining movements of said robot; and
wherein in said reprocessing phase said computer associates in said database actions said instructions to at least one of said scenes.