Granted Patent B2

US 12,521,875 · App. 17/644,945 · Granted Jan 13, 2026

Inventors: Andrew James Barry (Cambridge, MA); Alfred Anthony Rizzi (Waltham, MA)

Assignee: Boston Dynamics, Inc.

B25J9/163B25J9/1697B25J13/006

Abstract

Claims (67)

1 . A computer-implemented method when executed by data processing hardware of a robot causes the data processing hardware to perform operations comprising:

receiving a three-dimensional point cloud of initial sensor data for a space within an environment about the robot;

receiving a selection input indicating a user-selection of a target object represented in an image corresponding to the space, the target object for grasping by an end-effector of a robotic manipulator of the robot;

generating a grasp region for the end-effector of the robotic manipulator by projecting a plurality of rays from the selected target object of the image onto the three-dimensional point cloud of initial sensor data;

determining an initial grasp geometry for the robotic manipulator to grasp the target object within the grasp region based on an initial pose for the end-effector of the robotic manipulator;

instructing the end-effector of the robotic manipulator to grasp the target object within the grasp region based on the initial grasp geometry;

receiving updated sensor data for a second pose of the end-effector of the robotic manipulator in response to the end-effector of the robotic manipulator executing the initial grasp geometry, the updated sensor data including additional information about the target object indicating that a foreign object is present that obstructs the end-effector from grasping the target object based on the initial grasp geometry, the foreign object not included in the initial sensor data;

modifying the initial grasp geometry based on the additional information about the target object such that the foreign object does not obstruct the end-effector of the robotic manipulator from grasping the target object based on the modified initial grasp geometry; and

controlling the end-effector of the robotic manipulator to grasp the target object within the grasp region based on the modified initial grasp geometry.

2 . The method of claim 1 , wherein determining the initial grasp geometry for the robotic manipulator to grasp the target object within the grasp region comprises:

generating a plurality of candidate grasp geometries based on the target object within the grasp region;

for each candidate grasp geometry of the plurality of candidate grasp geometries, determining a grasping score for grasping the target object, the grasping score indicating a likelihood of success for grasping the target object using the end-effector of the robotic manipulator based on the respective candidate grasp geometry; and

selecting the respective candidate grasp geometry with a greatest grasping score as the initial grasp geometry designated to grasp the target object, the initial grasp geometry based on an initial pose for the end-effector of the robotic manipulator.

3 . The method of claim 2 , wherein the operations further comprise:

determining a new set of candidate grasping geometries based on the updated sensor data, each candidate grasping geometry of the new set of candidate grasping geometries comprising a respective grasping score; and

determining that a respective candidate grasp geometry from the new set of candidate grasping geometries comprises a corresponding grasping score that exceeds the grasping score of the initial grasp geometry;

wherein modifying the initial grasp geometry is further based on the respective candidate grasp geometry from the new set of candidate grasping geometries.

4 . The method of claim 1 , wherein the operations further comprise receiving an end-effector constraint constraining one or more degrees of freedom for the end-effector of the robotic manipulator to grasp the target object.

5 . The method of claim 1 , wherein the target object represented in the image corresponding to the space corresponds to an object classified by a machine learning algorithm.

6 . The method of claim 1 , wherein the operations further comprise:

receiving the image corresponding to the space; and

classifying graspable objects within the received image using a machine learning object classification algorithm.

7 . The method of claim 1 , wherein instructing the end-effector of the robotic manipulator to grasp the target object within the grasp region based on the modified initial grasp geometry comprises:

instructing a body of the robot to pitch toward the target object; or

instructing a first leg of the robot to rotate an upper member of the first leg about a knee joint towards a lower member of the first leg.

8 . The method of claim 1 , wherein receiving the selection input occurs at a user device in remote communication with the data processing hardware of the robot.

9 . The method of claim 1 , wherein the operations further comprise:

receiving the image corresponding to the space; and

calibrating the received image.

10 . The method of claim 1 , wherein the end-effector comprises a gripper having a movable jaw and a fixed jaw, the movable jaw configured to move relative to the fixed jaw to move between an open position for the gripper and a closed position for the gripper.

11 . The method of claim 1 , wherein the robotic manipulator comprises one or more sensors mounted at or near the end-effector to capture the initial sensor data defining the three-dimensional point cloud for the space within the environment about the robot.

12 . A robot comprising:

a body;

a plurality of legs coupled to the body;

a robotic manipulator coupled to the body, the robotic manipulator comprising an end-effector configured to grasp objects within an environment about the robot;

data processing hardware in communication with the robotic manipulator; and

memory hardware in communication with the data processing hardware, the memory hardware storing instructions that when executed on the data processing hardware cause the data processing hardware to perform operations comprising:

receiving a three-dimensional point cloud of initial sensor data for a space within the environment about the robot;

receiving, from a user of the robot, a selection input indicating a user-selection of a target object represented in an image corresponding to the space, the target object for grasping by the end-effector of the robotic manipulator;

generating a grasp region for the end-effector of the robotic manipulator by projecting a plurality of rays from the selected target object of the image onto the three-dimensional point cloud of initial sensor data;

determining an initial grasp geometry for the robotic manipulator to grasp the target object within the grasp region based on an initial pose for the end-effector of the robotic manipulator;

instructing the end-effector of the robotic manipulator to grasp the target object within the grasp region based on the initial grasp geometry;

receiving updated sensor data for a second pose of the end-effector of the robotic manipulator in response to the end-effector of the robotic manipulator executing the initial grasp geometry, the updated sensor data including additional information about the target object indicating that a foreign object is present that obstructs the end-effector from grasping the target object based on the initial grasp geometry, the foreign object not included in the initial sensor data;

modifying the initial grasp geometry based on the additional information about the target object such that the foreign object does not obstruct the end-effector of the robotic manipulator from grasping the target object based on the modified initial grasp geometry; and

controlling the end-effector of the robotic manipulator to grasp the target object within the grasp region based on the modified initial grasp geometry.

13 . The robot of claim 12 , wherein determining the initial grasp geometry for the robotic manipulator to grasp the target object within the grasp region comprises:

generating a plurality of candidate grasp geometries based on the target object within the grasp region;

for each candidate grasp geometry of the plurality of candidate grasp geometries, determining a grasping score for grasping the target object, the grasping score indicating a likelihood of success for grasping the target object using the end-effector of the robotic manipulator based on the respective candidate grasp geometry; and

selecting the respective candidate grasp geometry with a greatest grasping score as the initial grasp geometry designated to grasp the target object, the initial grasp geometry based on an initial pose for the end-effector of the robotic manipulator.

14 . The robot of claim 13 , wherein the operations further comprise:

determining a new set of candidate grasping geometries based on the updated sensor data, each candidate grasping geometry of the new set of candidate grasping geometries comprising a respective grasping score; and

determining that a respective candidate grasp geometry from the new set of candidate grasping geometries comprises a corresponding grasping score that exceeds the grasping score of the initial grasp geometry;

wherein modifying the initial grasping geometry is further based on the respective candidate grasp geometry from the new set of candidate grasp geometries.

15 . The robot of claim 12 , wherein the operations further comprise receiving an end-effector constraint constraining one or more degrees of freedom for the end-effector of the robotic manipulator to grasp the target object.

16 . The robot of claim 12 , wherein the target object represented in the image corresponding to the space corresponds to an object classified by a machine learning algorithm.

17 . The robot of claim 12 , wherein the operations further comprise:

receiving the image corresponding to the space; and

classifying graspable objects within the received image using a machine learning object classification algorithm.

18 . The robot of claim 12 , wherein instructing the end-effector of the robotic manipulator to grasp the target object within the grasp region based on the modified initial grasp geometry comprises:

instructing the body of the robot to pitch toward the target object; or

instructing a first leg of the plurality of legs of the robot to rotate an upper member of the first leg about a knee joint towards a lower member of the first leg.

19 . The robot of claim 12 , wherein receiving the selection input occurs at a user device in remote communication with the data processing hardware of the robot.

20 . The robot of claim 12 , wherein the operations further comprise:

receiving the image corresponding to the space; and

calibrating the received image.

21 . The robot of claim 12 , wherein the end-effector comprises a gripper having a movable jaw and a fixed jaw, the movable jaw configured to move relative to the fixed jaw to move between an open position for the gripper and a closed position for the gripper.

22 . The robot of claim 12 , wherein the robotic manipulator comprises one or more sensors mounted at or near the end-effector to capture the initial sensor data defining the three-dimensional point cloud for the space within the environment about the robot.

Assignments (1)

Continuity (2)

References Cited (118)