首页|'Manipulating Boxes Using A Zoned Gripper' in Patent Application Approval Proces s (USPTO 20240408746)
'Manipulating Boxes Using A Zoned Gripper' in Patent Application Approval Proces s (USPTO 20240408746)
扫码查看
点击上方二维码区域，可以放大扫码查看
原文链接
NETL
NSTL
This patent application is assigned to Boston Dynamics Inc. (Waltham, Massachuse tts, United States).The following quote was obtained by the news editors from the background informa tion supplied by theinventors: “Box-like objects represent a large percentage o f objects that need to be picked (i.e., removed from a pallet or holding contain er) in industrial, manufacturing, logistics, and commercial environments.Typica lly, box-like objects are characterized by at least one substantially planar pic king surface. Conventionally,during robotic picking, a picking robot handles kn own sizes, numbers, and types of boxes arrangedin a uniform manner on a structu red pallet. Using mechanical fixtures, some current systems pre-positiona palle t of boxes so that a robot can pick them from known pre-programmed locations. An y deviationfrom this known structure, either in the size of the box, the number of boxes, or the location of boxesresults in failure of the system. Unfortunat ely, computer-vision-based systems often rely on the boxeshaving clean edges at their boundaries and cannot accurately determine size and/or position of boxes thathave advertising, printed characters, printed logos, pictures, color, or an y other texture on them. Suchboxes have visual edges on their faces (i.e., edge s that do not correspond to an actual physical boundary ofthe box). Because cur rent computer-vision-based systems cannot distinguish the physical edges betweentwo different boxes from other visual edges on the faces of boxes, these system s tend to misjudge the sizeand position of the box(es). Problematically, pickin g and moving the box where the system has misjudgedits size and location may ei ther cause the box to slip from the grasp of the robot or may cause the robotto pick two or more boxes where it should have picked only one.”
Boston Dynamics Inc.BusinessEmerging TechnologiesMachine LearningRobotRobotics
NETL
NSTL
