Abstract: Systems and methods for measuring a fleet of self-driving vehicles are disclosed. The system comprises one or more self-driving vehicles, non-transitory computer-readable media in communication with the vehicles, a fleet-management system in communication with the media and vehicles, and a server in communication with the media. The fleet-management system is configured to store vehicle status records comprising a vehicle status pertaining to each of the one or more vehicles, and a timestamp in a vehicle status log on the media. The server has a processor that is configured to generate a fleet-performance report based on the vehicle status log.

Abstract: One variation of a method for tracking fresh produce in a store includes: accessing a first hyper-spectral image, of a produce display in a store, recorded at a first time; extracting a first spectral profile from a first region of the first hyper-spectral image depicting a first set of produce units in the produce display; identifying a first varietal of the first set of produce units; characterizing qualities (e.g., ripeness, bruising, spoilage) of the first set of produce units in the produce display based on the first spectral profile; and, in response to qualities of the first set of produce units in the produce display deviating from a target quality range assigned to the first varietal, generating a prompt to audit the first set of produce units in the produce display.

Abstract: System for guiding an instrument within a vascular network of a patient are disclosed. In some embodiments, the system receives a medical image from a medical imaging device and identifies a distal tip and a direction the instrument in the image. The system may then determine a waypoint for the distal tip of the instrument based at least in part on the position and direction of the distal tip of the instrument. The system may then generate a trajectory command for moving the instrument through the vascular network from the current position to the waypoint. The system may operate in a closed loop. The system may provide the trajectory command to a robotic medical system configured to move the instrument according to the command.

Abstract: Clearing landmines is a dangerous activity and the danger is increased when trying to clear a landmine without triggering an explosion. An excavator for landmines is provided, which includes a first actuator that vertically moves a mounting plate. Two or more secondary actuators are mounted to the mounting plate, and each of the secondary actuators are able to extend and retract a tool. Each respective tool has drill bits or prongs for penetrating a ground surface. A neutralizer is also provided, and it includes: a robotic clamp to grip a landmine and to reposition the same; a cutting tool to cut an opening into the landmine; a steaming module to eject steam into the opening; and a catcher to collect melted explosive material that exits the landmine.

Abstract: A dish cart includes a top with a slot for receiving a lift arm that may be lifted up through the slot to raise or lower dishes. Lobe openings emanating from the slot receive lobes of a lift paddle secured to the lift arm. The cart may define a gap below the top for permitting insertion of the lift paddle and the lift arm. The cart may be adjusted to receive differently sized dishes by means of rails of different sizes or adjustable vanes. A dish drop station includes the lift arm and lift paddle and a lift mechanism for raising and lower the lift arm. The dish drop station may include a cart present sensor and include a lock for retaining the cart. The controller may be programmed to the lower the lift arm to an unlock position that disengages the cart lock.

Abstract: Spatiotemporal robotic navigation may include providing a set of robots non-conflicting access to the same shared resources at different times so that the robots may operate without continually accounting for the locations of the other robots and workers operating in the particular site, without continually planning or updating paths after determining an initial path, and without continuously adjusting movements as the robots near one another. The spatiotemporal robotic navigation involves generating spatiotemporal plans. Each plan has a set of objectives that a robot is to execute by different time intervals. Each plan is generated so as to not conflict with the resources being accessed by other robots at time intervals set in the plans of other robots.

Abstract: Provided are systems and methods by which robots predictively detect objects that may obstruct robotic tasks prior to the robots performing those tasks. For instance, a robot may receive a task, and may obtain dimensions of a task object based on a first identifier obtained with the task or with a sensor of the robot. The robot may determine a first boundary for moving the task object based on the task object dimensions and an added buffer of space that accounts for imprecise robotic operation. The robot may detect a second identifier of a neighboring object, and may obtain dimensions of the neighboring object using the second identifier. The robot may compute a second boundary of the neighboring object based on the dimensions of the neighboring object and a position of the second identifier, and may detect an obstruction based on the second boundary crossing into the first boundary.

Abstract: This disclosure provides improved nautical craft that can travel and navigate on their own. A hybrid vessel is described that converts wave motion to locomotive thrust by mechanical means, and also converts wave motion to electrical power for storage in a battery. The electrical power can then be tapped to provide locomotive power during periods where wave motion is inadequate and during deployment. The electrical power can also be tapped to even out the undulating thrust that is created when locomotion of the vessel is powered by wave motion alone.

Abstract: In one embodiment, a system and method is provided for optimizing the total time taken for cleaning an area by a cleaning robot, which is the sum of the cleaning time and the time taken for the robot to re-charge its battery mid-cleaning when the area is too large to clean on a single charge. In one embodiment, the remaining area to be cleaned is determined, and the mid-cleaning re-charge time is reduced to the amount of charge needed to clean the remaining area, plus a buffer. Hence, a reduction in mid-cleaning charging time results in reduced total time taken to clean a given space.

Abstract: An embodiment provides a method of controlling a pipe inspection robot, including: detecting, from a headset, brain activity of a user wearing the headset; binning the brain activity of the user into one of a plurality of different bins; the binning comprising accumulating brain activity values for a period of time to establish a brain activity value for the period of time; determining, using the brain activity value, if a threshold level of brain activity has accumulated over the period of time; after the threshold level has been accumulated, identifying a control action to be sent to the pipe inspection robot based on the brain activity value; and controlling the movement of the pipe inspection robot using a control signal associated with the control action. Other embodiments are described and claimed.

Abstract: A robotic kitchen assistant for frying includes a robotic arm, a fryer basket, and a robotic arm adapter assembly allowing the robotic arm to pick up and manipulate the fryer basket. The robotic arm adapter includes opposing gripping members to engage the fryer basket. A utensil adapter assembly is mounted to the handle of the fryer basket, and the opposing gripper members are actuated to capture a three-dimensional (3D) feature of the utensil adapter assembly. The robotic arm adapter assembly can include an agitator mechanism to shake the fryer basket or another utensil as desired. Related methods are also described.

Abstract: Training images can be synthesized in order to obtain enough data to train a model (e.g., a neural network) to recognize various classifications of a type of object. Images can be synthesized by blending images of objects labeled using those classifications into selected background images. To improve results, one or more operations are performed to determine whether the synthesized images can still be used as training data, such as by verifying one or more objects of interested represented in those images is not occluded, or at least satisfies a threshold level of acceptance. The training images can be used with real world images to train the model.

Abstract: A method and an apparatus for controlling a robot are provided. According to one aspect of the present disclosure, the method can include receiving information on a work type of robot motion performed by the robot; generating workflow of the robot motion based upon the received information on the work type of the robot motion; measuring information on work environment in which the robot motion is performed in accordance with the work type and the workflow of the robot motion by controlling the robot; receiving work information on the robot motion in accordance with the work type and the workflow of the robot motion by controlling the robot; and performing the robot motion in accordance with the work type and the workflow of the robot motion by controlling the robot based upon the measured information on the work environment and the received work information on the robot motion.

Abstract: Provided are robots that autonomously detect and correct individualized anomalies resulting from deviations in the sensors and/or actuators of individual robots, and environmental anomalies resulting from deviations in the environment elements that the robots rely on or use in the execution of different tasks. To do so, a robot may receive a task, may determine expected kinematics that include expected activations of a set of sensors and actuators by which the robot executes the task, may activate the set of sensors and actuators according to the expected kinematics, may track the actual kinematics resulting from activating the set of sensors and actuators according to the expected kinematics and continuing the activations until detecting one or more environment elements signaling completion of the task, and may adjust one or more sensors, actuators, or environment elements in response to the actual kinematics deviating from the expected kinematics.

Abstract: A propulsion rudder for use with a water vessel, preferably an unmanned surface water vessel. The propulsion rudder includes a rotating base, a rudder assembly, and a thrust assembly. In one embodiment, the rudder assembly may comprise two vertical mast rudders connected by an upper horizontal wing, with the thrust assembly mounted on the horizontal wing. In a second embodiment, the rudder assembly may comprise an annular airfoil with the thrust assembly mounted on a mast and positioned in the center of the annular airfoil. The thrust assembly may be a propeller, a turbine, a ramjet, or a rocket.

Abstract: A controller is provided to orchestrate operations of two or more different actors or resources in completing a task. The controller orchestrates the operations based on a dependency graph that the controller generates for each received task. The dependency graph has a linked list of nodes. Each node includes a different set of operations for completing a different part of a task represented by that dependency graph. Each set of operations associated with a node may be dependent on the successful completion of a prior set of operations associated with a prior linked node in the same dependency graph. The controller may allocate different sets of operations from different nodes of the dependency graph to different human or robotic resources, and may track the status of multiple tasks in parallel based on the progression through different nodes of the corresponding dependency graphs.

Abstract: Exemplary embodiments relate to user-assisted robotic control systems, user interfaces for remote control of robotic systems, vision systems in robotic control systems, and modular grippers for use by robotic systems. Systems, methods, apparatuses and computer-readable media instructions are disclosed for interactions with and control of robotic systems, in particular, pick and place systems using soft robotic actuators to grasp, move and release target objects.

Abstract: An actuator and end effector are controlled according to images from cameras having a surface in their field of view. Vessels (cups, bowls, etc.) and other objects are identified in the images and their configuration is assigned to a finite set of categories by a classifier that does not output a 3D bounding box or determine a 6D pose. For objects assigned to a first subset of categories, grasping parameters for controlling the actuator and end effector are determined using only 2D bounding boxes, such as oriented 2D bounding boxes. For objects not assigned to the first subset, a righting operation may be performed using only 2D bounding boxes. Objects that are still not in the first set may then be grasped by estimating a 3D bounding box and 6D pose.

Abstract: Systems and methods for safety-enabled control.

Abstract: A virtual put wall system includes a storage apparatus with multiple containers, and different container identifiers attached to the containers. The system includes databases storing a first mapping that maps each customer order of a plurality of customer orders to a different container, a second mapping that maps each container to a different container identifier attached to that container, and a different visual identifier of each container. The system also includes a display device and a coordinating device that detects retrieval of an object, determines that the object belongs to a particular customer order, selects, based on the second mapping, a particular container that is used to store objects of the particular customer order, and directs the transfer of the object to the particular container by modifying the display device to present a particular visual identifier of the particular container.