US 11,534,923 B1Granted

Robotic manipulation and frame sensor placement for in-motion object modeling and identification

AssigneeAmazon Technologies, Inc.

InventorFelipe De Arruda Camargo Polido

Life of this patent

Estimating the term…

Loading patent…

Abstract

A package imaging system can be configured to utilize unused or idle degrees of freedom associated with a robot to collect identification and modeling information. One or more sensors can be attached to the robot and configured to track a package between a first location and a second location. A final joint of the robot can be configured to cause one or more rotations of the package within the reference frame. The one or more sensors can capture one or more images of the package within the reference frame. Based on the one or more images, the package imaging system can determine a package identity from an identifier associated with the package. The package imaging system can also generate a three-dimensional model of the package by combining the one or more pictures.

Claims20

Dependent claims

A method comprising:

monitoring, via one or more sensors associated with a robot, a reference frame that occupies a fixed location relative to a reference joint of the robot;
causing the robot to transfer a package from a first location to a second location;
determining that the package is within the reference frame;
causing, based at least in part on the package being within the reference frame, at least a rotation of the package within the reference frame;
causing the one or more sensors to capture one or more images of the package during the rotation of the package;
and
determining a package identity of the package based at least in part on the one or more images by:
generating, based at least in part on the one or more images, dimension information associated with the package;
determining that the dimension information is unique to a package profile associated with the first location;
and
determining, based at least in part on the package profile, the package identity for the package.

8 more elements

A system comprising:

one or more processors;
and
a memory storing one or more instructions that are executable by the one or more processors to perform operations comprising:
monitoring, via one or more sensors associated with a robot, a reference frame that occupies a fixed location relative to a reference joint of the robot;
causing the robot to transfer a package from a first location to a second location;
determining that the package is within the reference frame;
causing, based at least in part on the package being within the reference frame, at least a rotation of the package within the reference frame;
causing the one or more sensors to capture one or more images of the package during the rotation of the package;
and
determining a package identity of the package based at least in part on the one or more images by:
generating, based at least in part on the one or more images, dimension information associated with the package;
determining that the dimension information is unique to a package profile associated with the first location;
and
determining, based at least in part on the package profile, the package identity for the package.

11 more elements

A system comprising:

a robot comprised of a first joint and a final joint that are connected via one or more connecting joints, a vacuum nozzle or a grasping mechanism of the robot configured to remove a package from a storage container at a first physical location and deposit the package onto a conveyor belt at a second physical location that is different than the first physical location, the final joint configured to rotate the package in a direction while the robot is moving the package from the storage container to the conveyor belt;
a first sensor attached to the first joint and configured to detect packages within a volume of space from a first perspective;
a second sensor attached to the first joint and configured to detect the packages within the volume of space from a second perspective that is different than the first perspective;
and
a control system configured to:
determine that the final joint and the package are within the volume of space;
cause the final joint to complete one or more rotations of the package within the volume of space as the robot moves the package from the storage container to the conveyor belt;
receive, from the first sensor and the second sensor, multiple images that depict the package within the volume of space in different orientations;
identify, based on the multiple images, a barcode on an exterior surface of the package;
determine, based on the multiple images, dimensions of the package;
and
determine, based on at least one of the barcode or the dimensions, an identity of the package.

9 more elements

Dependencies are inferred from the claim text.