Abstract: Systems and techniques for performing robotic picking of items based on a reference image and item attributes are described. In an example technique, an image of a picking environment having multiple items is obtained. A reference image of a first item is obtained, based on an item ID associated with the first item. An instance segmentation operation is performed on the image of the picking environment to determine an instance of the first item within the image of the picking environment, based on the reference image and the image of the picking environment. A grasp plan for a robotic arm end-of-arm tool is generated based on segmentation information output from the instance segmentation operation and the item attributes. The end-of-arm tool is controlled to pick an item from the multiple items according to the grasp plan.

Abstract: Autonomous robotic pack planning systems and methods may include an item identification subsystem, a pack planning subsystem, and a robotic packing subsystem. Based on identified items to be packed, the pack planning subsystem may generate a pack plan with placement poses for items that satisfy various constraints, including containment constraints, stability constraints, and/or stress constraints. Then, the robotic packing subsystem may pack the items according to the pack plan. Further, packing of items may be monitored in real-time, and the pack plan may be dynamically adjusted or regenerated in response to deviations from the pack plan.

Abstract: A computing system assigns items in a received purchase order to containers and coordinates movement of a first set of robots holding the containers and movement of a second set of robots holding the items to a set of stations that include robotic arms. The computing system also uses the robotic arms to place the items into the containers.

Abstract: A gripping tool includes first and second fingers coupled to a base member such that a distance between the first and second fingers is adjustable. First and second grippers are attached to the first and second fingers, respectively. The first and second grippers each have a compliant gripping surface and are each selectively expandable so as to reduce the distance and grip select portions of an item. A first actuator is operatively coupled to the first gripper and is configured to selectively expand the first gripper in a first direction. A second actuator is operatively coupled to the second gripper and is configured to selectively expand the second gripper in a second direction. At least one of the first and second actuators is configured to vary an expansion force of the respective first gripper in the first direction or the second gripper in the second direction.

Abstract: Systems, methods, and computer-readable media are disclosed for automated robotic sortation. In one embodiment, an example system may include a first robotic arm configured to grasp a moving object from a conveyor while the conveyor is in motion, a camera system configured to image objects moving on the conveyor, and a first scanner configured to scan machine-readable codes on objects moving on the conveyor. The system may be configured to sort objects from the conveyor using the first robotic arm based at least in part on an object destination.

Abstract: Systems, methods, and computer-readable media are disclosed for automated custom shipping containers. In one embodiment, an example system may include a depth camera, a first robotic arm configured to grasp objects, a box making machine configured to generate boxes, and a computer system. The computer system may be configured to determine a depth map of a first object and a second object using the depth camera, determine a first dimension of the first object, determine a second dimension of the second object, and determine a first box dimension for a box in which the first object and the second object are to be placed using the first dimension and the second dimension. The computer system may be configured to cause the box making machine to produce the box, and cause the first robotic arm to place the first object and the second object into the box.

Abstract: Systems, methods, and computer-readable media are disclosed for automated robotic sortation. In one embodiment, an example system may include a first robotic arm configured to grasp a moving object from a conveyor while the conveyor is in motion, a camera system configured to image objects moving on the conveyor, and a first scanner configured to scan machine-readable codes on objects moving on the conveyor. The system may be configured to sort objects from the conveyor using the first robotic arm based at least in part on an object destination.

Abstract: Techniques for identifying optimal placement instructions for placing an object within a container by a robotic device are disclosed. An artificial intelligence model may be trained utilizing historical placement images and/or placement data to identify subsequent placement instructions for a robotic device. As an object is brought within a threshold distance of a goal placement location, one or more images may be captured of an area adjacent to the placement location. Features of these images may be provided to the artificial intelligence model as input to cause the artificial intelligence model to identify particular placement instructions for the robotic device. By executing the identified placement instructions, the available space within the storage container may be optimized (e.g., increased) by reducing the space between objects of the storage container in accordance with the placement instructions.

Abstract: Systems and methods are provided herein for utilizing multi-functional storage containers to store items within a facility. An indication may be received that an item is to be stored within some storage container within the facility. The storage container may be a multi-functional storage container configured with a first set of storage compartments designated for items that are to be manually placed and a second set of storage compartments designated for items that are to be robotically placed. A placement method (e.g., robotic or manual) may be determined for the item and a storage component of the storage container selected based at least in part on the determined placement method. Once the storage component is selected, instructions may be provided to a robotic device (for robotic placement) or a computing device (for manual placement) instructing that the item be placed at the selected storage component.

Abstract: Embodiments herein describe a robotic pick and stow storage system with a built-in consolidation function that increases the storage density. The system can include a stowing robot, a consolidation robot, and individual containers. The stowing robot may pick an item, identify a free space in the container large enough for the item, and then place the item in the free space. The items placed in the container may have a space between it and a neighboring item or items (referred to herein as a stow buffer) due to uncertainties or inaccuracies in the stowing process. The embodiments herein describe a stow and consolidation action that reduces the stow buffers thereby providing additional space in the container for more items. The consolidation robot can tilt and vibrate the container such that the items move closer together thereby generating free space for additional items in the container.