Abstract: Methods, apparatus, and computer readable storage media related to defining a planar model that approximates a plurality of surfaces of an object and/or applying the planar model to detect the object and/or to estimate a pose for the object. For example, the planar model may be compared to data points sensed by the a three-dimensional vision sensor of a robot to determine the object is present in the field of view of the sensor and/or to determine a pose for the object relative to the robot. A planar model comprises a plurality of planar shapes modeled relative to one another in a three-dimensional space and approximates an object by approximating one or more surfaces of the object with the planar shapes.

Abstract: Efficient energy storage is provided by using a working fluid flowing in a closed cycle including a ganged compressor and turbine, and capable of efficient heat exchange with heat storage fluids on a hot side of the system and on a cold side of the system. This system can operate as a heat engine by transferring heat from the hot side to the cold side to mechanically drive the turbine. The system can also operate as a refrigerator by mechanically driving the compressor to transfer heat from the cold side to the hot side. Heat exchange between the working fluid of the system and the heat storage fluids occurs in counter-flow heat exchangers. In a preferred approach, molten salt is the hot side heat storage fluid and water is the cold side heat storage fluid.

Abstract: Example pallet-conveyor systems may include a conveyor system configured with a delivery track arranged to move pallets to a delivery area, a recirculation loop, and a diverter mechanism. The system may include a computing system that selects an item for the recirculation loop based on future demand and causes a robotic device to maintain pallets of the selected item in the recirculation loop. The computing system may further receive an item request and determine that a requested item is available from the recirculation loop and responsively cause the diverter mechanism to divert the requested item from the recirculation loop to the delivery track. The computing system may also cause robotic devices to obtain and load pallets of remaining requested items onto the conveyer system for the delivery area, and cause pickers to remove one or more items from pallets at the delivery area in order to fulfill the item request.

Abstract: Methods and apparatus related to training and/or utilizing a convolutional neural network to generate grasping parameters for an object. The grasping parameters can be used by a robot control system to enable the robot control system to position a robot grasping end effector to grasp the object. The trained convolutional neural network provides a direct regression from image data to grasping parameters. For example, the convolutional neural network may be trained to enable generation of grasping parameters in a single regression through the convolutional neural network. In some implementations, the grasping parameters may define at least: a “reference point” for positioning the grasping end effector for the grasp; and an orientation of the grasping end effector for the grasp.

Abstract: The present disclosure relates to systems, devices, and methods for calibrating a light field projection system. One example system includes a projection unit operable to project a scanning sequence toward a screen having convex reflective elements. The scanning sequence is modulated according to a baseline intensity profile. The system also includes a calibration device disposed such that a portion of the scanning sequence is intercepted by the calibration device. The calibration device includes a first light detector arranged to detect an intercepted intensity profile. The calibration device also includes a second light detector arranged to detect a reflected portion of the scanning sequence as a measured intensity profile. The system further includes a control system. The control system is configured to determine an expected intensity profile and to modify operation of the light field projection system based on a comparison of the measured intensity profile to the expected intensity profile.

Abstract: An assembly for manufacturing a balloon envelope includes a table component and a sealing component. The table component may include a first platform, a second platform, a third platform, and a lateral opening between the first and second platforms. The first and second platforms can receive a first sheet of material that forms a first gore of the balloon envelope and a second sheet of material that forms a second gore of the balloon envelope. At least a portion of the first platform may move relative to the third platform so as to allow for the tendon and the portion of the second sheet attached to the tendon to move toward the third platform. The sealing component may be configured to bond the first sheet to the second sheet in order to join the first and the second gores of the balloon envelope.

Abstract: While motors or generators stacked in series may allow for higher operating voltages, such motors or generators may also exhibit instability. To minimize instability, the motors or generators may be controlled to have an approximately equal current. An example motor system may include motor stacks connected in series, each motor stack exhibiting a respective stack voltage and a respective differential power (based on a difference in power between motors in the motor stack). A control system may average the stack voltages to generate an average stack voltage and generate a nominal stack power corresponding to each stack voltage. The control system may receive the differential powers, combine each differential power and nominal stack power for the respective motor stack to generate first and a second motor powers, and control each motor stack using the first and second motor powers.

Abstract: Examples described may enable consolidating pallets of items in a warehouse. An example method includes receiving real-time item information including pallet locations in a warehouse and inventory of items arranged on the pallets; based on the real-time item information, identifying a set of pallets of which at least one pallet includes less than a threshold quantity of a type of item; receiving real-time robotics information and determining, based on the real-time item and robotics information, an amount of time to condense the items on the set of pallets into a single pallet and a quantity of pallets that will become empty as a result of condensing the items; and, based on the amount of time being less than a threshold time and the quantity of pallets exceeding a threshold quantity of pallets, causing robotic devices to condense the items into the single pallet.

Abstract: An example system includes a transmission having a first plurality of gears and an extent of backlash. The system also includes a first motor connected to an input shaft of the transmission and a second motor connected to an output shaft of the transmission through a second plurality of gears. A first gear ratio of a first plurality of gears is greater than a second gear ratio of the second plurality of gears. The system may receive a command to change a direction of rotation of the output shaft from a first direction to a second direction. In response to the received command, the first motor may drive the transmission through a first portion of the extent of backlash deadband. The second motor may drive the transmission through a second portion of the extent of backlash deadband. The second portion may be greater than the first portion.

Abstract: Systems and methods are provided for specifying safety rules for robotic devices. A computing device can determine information about any actors present within a predetermined area of an environment. The computing device can determine a safety classification for the predetermined area based on the information. The safety classification can include: a low safety classification if the information indicates zero actors are present within the predetermined area, a medium safety classification if the information indicates any actors are present within the predetermined area all are of a predetermined first type, and a high safety classification if the information indicates at least one actor present within the predetermined area is of a predetermined second type. After determining the safety classification for the predetermined area, the computing device can provide a safety rule for operating within the predetermined area to a robotic device operating in the environment.

Abstract: A heat engine system with pressure-regulating load-locks disposed between thermal medium storage containers and heat exchangers is disclosed. A load-lock connects one or more storage containers at atmospheric pressure to one or more heat exchangers at greater than or less than atmospheric pressure.

Abstract: Closed thermodynamic cycle systems, such as closed Brayton cycle systems, with regenerative heat exchangers are disclosed. Embodiments include dual regenerators and regenerators with buffer tank systems. Regenerators may be used instead of or in addition to one or more recuperators within the systems, and may be used as a means of gas-gas heat exchange for different streams of a working fluid.

Abstract: Methods, apparatus, systems, and computer-readable media are provided for efficient mapping of a robot environment. In various implementations, a group of data points may be sensed by a three-dimensional sensor. One or more voxels of a three-dimensional voxel model that are occupied by the group of data points may be identified. For each occupied voxel, a column of the three-dimensional voxel model that contains the occupied voxel may be identified. Occupied voxels contained in each column may be indexed by elevation. In various implementations, one or more sparse linked data structures may be used to represent the columns.

Abstract: An example embodiment may involve receiving a request to provide unmanned aerial vehicle (UAV) based wireless coverage to a particular geographical location. Possibly in response to the request, a UAV may fly to the particular geographical location. A first wireless interface of the UAV may define a wireless coverage area that covers at least part of the particular geographical location. A second wireless interface of the UAV may establish a wireless backhaul link to a data network. The UAV may provide wireless data transfer services to at least one device in the particular geographical location, where the wireless data transfer services allow the device to exchange data communication with the data network via the UAV.

Abstract: An example system includes a robotic device deployed in a warehouse environment including a plurality of inventory items. The system also includes a camera coupled to the robotic device, configured to capture image data. The system also includes a computing system configured to receive the captured image data. The computing system is configured to, based on the received image data, generate a navigation instruction for navigation of the robotic device. The computing system is also configured to analyze the received image data to detect one or more on-item visual identifiers corresponding to one or more inventory items. The computing system is further configured to, for each detected visual identifier, (i) determine a warehouse location of the corresponding inventory item, (ii) compare the determined warehouse location to an expected location, and (iii) initiate an action based on the comparison.

Abstract: A free space optical communication system transmits and receives optical signals in a colorless manner using an optical circulator. The system installs the optical circulator with a single mode (SM) fiber at port 1, a double clad (DC) fiber at port 2, and a multimode (MM) fiber at port 3. The system injects a first optical signal into a core of the SM fiber. The system then routes the first optical signal at port 1, using the optical circulator, into a SM core of the DC fiber via Port 2. Further, the system injects a second optical signal into a first cladding of the DC fiber. The system then routes the second optical signal at port 2, using the optical circulator, into the MM fiber via Port 3.

Abstract: Methods, apparatus, systems, and computer-readable media are provided for avoiding and/or operating robots through singularities. In various implementations, a Cartesian input velocity to be attained by a robot end effector may be received as user input. An attainable velocity of the end effector may be determined that excludes at least one directional component deemed unattainable by the end effector based on physical attribute(s) of the robot. Joint velocities to be attained by joints of the robot in joint space to move the end effector pursuant to the attainable velocity in configuration space may be calculated and scaled to account for a joint velocity limit associated with at least one joint. Actuator trajectories may be calculated for the joints of the robot based at least in part on the scaled joint velocities. The joints of the robot may then be operated in accordance with the calculated joint trajectories.

Abstract: An example method includes receiving instructions to pick up an object with one or more lift elements of an autonomous vehicle. Based on a current positioning of the vehicle, the method further includes identifying the object to be picked up and a particular side of the object under which to place the one or more lift elements of the vehicle. The method additionally includes determining an approach path toward the object for the vehicle to follow to place the lift elements of the vehicle under the particular side of the object. The method further includes causing the vehicle to move along the determined approach path toward the object. The method additionally includes determining that the lift elements of the vehicle are placed under the particular side of the object. The method also includes causing the vehicle to lift the object with the lift elements.

Abstract: A 3D printing process may form a 3D object by alternatingly forming layers from a liquid resin and a solid. For instance, when printing a 3D object, the 3D printer may at least partially cure a layer of liquid resin, and before the curing of the resin is complete, dip the semi-cured resin into a vat containing graphene powder so as to create a super strong 3D object. As another example, each semi-cured resin layer could be pressed into a vat of fiberglass such that the fiberglass is coupled to the semi-cured resin. The resin may then be allowed to finish curing before the next layer of resin is formed. In other embodiments, this process could be used to embed sensors in 3D printed objects.

Abstract: Methods and systems for providing landmarks to facilitate robot localization and visual odometry are provided herein. At least one area of a physical environment in which a robotic device resides may be determined to include surfaces that lack sufficient discernable features to determine a location of the at least one area. Instructions may responsively be provided to the robotic device for the robotic device to provide a material in respective patterns onto one or more surfaces of the at least one area. Instructions can responsively be provided for the robotic device to provide the material in respective textures as well. The respective patterns or textures may include sufficient discernable features to determine a location of the at least one area, and the material may remain on the one or more surfaces for a predetermined period of time.