Abstract: A method of manipulating boxes includes receiving a minimum box size for a plurality of boxes varying in size located in a walled container. The method also includes dividing a grip area of a gripper into a plurality of zones. The method further includes locating a set of candidate boxes based on an image from a visual sensor. For each zone, the method additionally includes, determining an overlap of a respective zone with one or more neighboring boxes to the set of candidate boxes. The method also includes determining a grasp pose for a target candidate box that avoids one or more walls of the walled container. The method further includes executing the grasp pose to lift the target candidate box by the gripper where the gripper activates each zone of the plurality of zones that does not overlap a respective neighboring box to the target candidate box.

Abstract: A method of manipulating boxes includes receiving a minimum box size for a plurality of boxes varying in size located in a walled container. The method also includes dividing a grip area of a gripper into a plurality of zones. The method further includes locating a set of candidate boxes based on an image from a visual sensor. For each zone, the method additionally includes, determining an overlap of a respective zone with one or more neighboring boxes to the set of candidate boxes. The method also includes determining a grasp pose for a target candidate box that avoids one or more walls of the walled container. The method further includes executing the grasp pose to lift the target candidate box by the gripper where the gripper activates each zone of the plurality of zones that does not overlap a respective neighboring box to the target candidate box.

Abstract: Method and apparatus for object detection by a robot are provided. The method comprises analyzing using a set of trained detection models, one or more first images of an environment of the robot to detect one or more objects in the environment of the robot, generating at least one fine-tuned model by training one or more of the trained detection models in the set, wherein the training is based on a second image of the environment of the robot and annotations associated with the second image, wherein the annotations identify one or more objects in the second image, updating the set of trained detection models to include the generated at least one fine-tuned model, and analyzing using the updated set of trained detection models, one or more third images of the environment of the robot to detect one or more objects in the environment.

Abstract: A method of manipulating boxes includes receiving a minimum box size for a plurality of boxes varying in size located in a walled container. The method also includes dividing a grip area of a gripper into a plurality of zones. The method further includes locating a set of candidate boxes based on an image from a visual sensor. For each zone, the method additionally includes, determining an overlap of a respective zone with one or more neighboring boxes to the set of candidate boxes. The method also includes determining a grasp pose for a target candidate box that avoids one or more walls of the walled container. The method further includes executing the grasp pose to lift the target candidate box by the gripper where the gripper activates each zone of the plurality of zones that does not overlap a respective neighboring box to the target candidate box.

Abstract: A method of manipulating boxes includes receiving a minimum box size for a plurality of boxes varying in size located in a walled container. The method also includes dividing a grip area of a gripper into a plurality of zones. The method further includes locating a set of candidate boxes based on an image from a visual sensor. For each zone, the method additionally includes, determining an overlap of a respective zone with one or more neighboring boxes to the set of candidate boxes. The method also includes determining a grasp pose for a target candidate box that avoids one or more walls of the walled container. The method further includes executing the grasp pose to lift the target candidate box by the gripper where the gripper activates each zone of the plurality of zones that does not overlap a respective neighboring box to the target candidate box.

Abstract: A method of manipulating boxes includes receiving a minimum box size for a plurality of boxes varying in size located in a walled container. The method also includes dividing a grip area of a gripper into a plurality of zones. The method further includes locating a set of candidate boxes based on an image from a visual sensor. For each zone, the method additionally includes, determining an overlap of a respective zone with one or more neighboring boxes to the set of candidate boxes. The method also includes determining a grasp pose for a target candidate box that avoids one or more walls of the walled container. The method further includes executing the grasp pose to lift the target candidate box by the gripper where the gripper activates each zone of the plurality of zones that does not overlap a respective neighboring box to the target candidate box.

Abstract: Techniques are described that enable robotic picking systems to locate and pick boxes from an unstructured pallet using computer vision and/or one or more “exploratory picks” to determine the sizes and locations of boxes on the pallet.

Abstract: Techniques are described that enable robotic picking systems to locate and pick boxes from an unstructured pallet using computer vision and/or one or more “exploratory picks” to determine the sizes and locations of boxes on the pallet.