Abstract: In some embodiments, an unmanned aerial vehicle (UAV) is provided. The UAV comprises one or more processors; a camera; one or more propulsion devices; and a computer-readable medium having instructions stored thereon that, in response to execution by the one or more processors, cause the UAV to perform actions comprising: receiving at least one image captured by the camera; generating labels for pixels of the at least one image by providing the at least one image as input to a machine learning model; identifying one or more landing spaces in the at least one image based on the labels; determining a relative position of the UAV with respect to the one or more landing spaces; and transmitting signals to the one or more propulsion devices based on the relative position of the UAV with respect to the one or more landing spaces.

Abstract: Systems and methods for validating a position of an unmanned aerial vehicle (UAV) are provided. A method can include receiving map data for a location, the map data including labeled data for a plurality of landmarks in a vicinity of the location. The method can include generating image data for the location, the image data being derived from images of the vicinity generated by the UAV including at least a subset of the plurality of landmarks. The method can include determining a visual position of the UAV using the image data and the map data. The method can include determining a Global Navigation Satellite System (GNSS) position of the UAV. The method can include generating an error signal using the visual position and the GNSS position. The method can also include validating the GNSS position in accordance with the error signal satisfying a transition condition.

Abstract: In some embodiments, a method for autonomous navigation of an unmanned aerial vehicle (UAV) is provided. The UAV determines a tracked position using at least one positioning sensor of the UAV. The UAV captures an image using a camera of the UAV. The UAV determines a visual position confidence area using the captured image. The UAV checks the tracked position using the visual position confidence area to determine whether the tracked position is accurate. In response to determining that the tracked position is not accurate, the UAV causes a corrective action based on the visual position confidence area to be taken.

Abstract: An uncrewed aerial vehicle (UAV) may be configured to hover above a particular charging pad within a portion of a cluster of charging pads for UAVs. The cluster may include the charging pads arranged in a layout and fiducial markers distributed at positions across the layout. While hovering above the particular charging pad, the UAV may capture an aerial image of the portion of the cluster. The UAV may derive cluster-portion observation data from the image, the cluster-portion observation data including information indicating a position of the particular charging pad, and positions of one or more fiducial markers within the portion of the cluster relative to the particular charging pad. The UAV may send the cluster-portion observation data to a computing system in an infrastructure support network for UAVs, and thereafter receive, from the computing system, location information indicating that UAV's geolocation is a geolocation of the particular charging pad.

Abstract: In some embodiments, a method of determining an estimated location of a UAV is provided. A captured image is received from a camera of the UAV. Semantic labels are generated by the UAV for a plurality of objects visible in the captured image. The UAV compares the semantic labels to reference labels associated with a reference map to determine a current location estimate. The UAV updates an accumulated location estimate using the current location estimate, and the UAV determines the estimated location of the UAV based on the accumulated location estimate.

Abstract: In some embodiments, an unmanned aerial vehicle (UAV) is provided. The UAV comprises one or more processors; a camera; one or more propulsion devices; and a computer-readable medium having instructions stored thereon that, in response to execution by the one or more processors, cause the UAV to perform actions comprising: receiving at least one image captured by the camera; generating labels for pixels of the at least one image by providing the at least one image as input to a machine learning model; identifying one or more landing spaces in the at least one image based on the labels; determining a relative position of the UAV with respect to the one or more landing spaces; and transmitting signals to the one or more propulsion devices based on the relative position of the UAV with respect to the one or more landing spaces.

Abstract: Systems and methods for validating a position of an unmanned aerial vehicle (UAV) are provided. A method can include receiving map data for a location, the map data including labeled data for a plurality of landmarks in a vicinity of the location. The method can include generating image data for the location, the image data being derived from images of the vicinity generated by the UAV including at least a subset of the plurality of landmarks. The method can include determining a visual position of the UAV using the image data and the map data. The method can include determining a Global Navigation Satellite System (GNSS) position of the UAV. The method can include generating an error signal using the visual position and the GNSS position. The method can also include validating the GNSS position in accordance with the error signal satisfying a transition condition.

Abstract: A computer-implemented method comprises receiving an image captured by a camera on an unmanned aerial vehicle (UAV). The image depicts an environment below the UAV. A feature mask associated with the image is generated via a machine learning model that is trained to identify and semantically label pixels representing the environment depicted in the image. One or more reference tiles associated with the environment are retrieved. The reference tiles are associated with particular geographic locations and specify semantically labeled pixels representing the geographic locations. The semantically labeled pixels of the feature mask are correlated with the semantically labeled pixels of at least one of the one or more reference tiles to determine the geographic location of the UAV in the environment.

Abstract: Aspects of the present disclosure involve a system and a method for performing operations comprising: storing, on a distributed storage system, a front-end (FE) instance and a plurality of real-time graph (RTG) instances, each of the plurality of RTG instances includes a plurality of device objects, the FE instance being configured to communicate with a client device associated with a first user; establishing a bi-directional streaming remote procedure call (RPC) connection between the FE instance and the plurality of RTG instances; receiving, by the FE instance, a status update from the client device; determining, by the FE instance, that a first device object corresponding to the client device is stored on a first RTG instance of the plurality of RTG instances; and transmitting a first message comprising the status update from the FE instance to the first RTG instance to update the first device object.

Abstract: Aspects of the present disclosure involve a system and a method for performing operations comprising: storing, on a distributed storage system, a plurality of real-time graph (RTG) instances that include a plurality of device objects; receiving, by a first device object of the plurality of device objects, a status update from a client device associated with a first user; transmitting, by the first device object, a first message comprising the status update to a second device object associated with a second user over a real-time link; and transmitting, by the first device object, a second message comprising the status update to a third device object associated with a third user over a first rate-limited link.

Abstract: Aspects of the present disclosure involve a system and a method for performing operations comprising: storing, on a distributed storage system, a plurality of real-time graph (RTG) instances that include a plurality of device objects, the plurality of device objects comprising a first device object associated with a given user; receiving, by the first device object, a friends list of the given user having a first version identifier; receiving, by the first device object, an update from a client device associated with the given user, the update comprising a friends list version identifier; determining that the first version identifier of the friends list in the first device object mismatches the friends list version identifier in the update; and synchronizing the friends list in the first device object prior to sending one or more messages that include the update to other device objects of the plurality of device objects.

Abstract: Aspects of the present disclosure involve systems and methods for performing operations comprising: storing, on a distributed storage system, one or more front-end (FE) instances and a plurality of real-time graph (RTG) instances; receiving, by a first online device object associated with a given user, an update from a first client device; generating, by the first online device object, a message that includes the update for transmission to a plurality of friends of the given user; storing, on a given one of the plurality of RTG instances, an offline device object for a first friend of the plurality of friends and a second online device object for a second friend of the plurality of friends; and transmitting, by the first online device object, the message that includes the update to the offline device object of the first friend and the second online device object of the second friend.

Abstract: Aspects of the present disclosure involve a system and a method for performing operations comprising: storing, on a distributed storage system, a plurality of real-time graph (RTG) instances that include a plurality of device objects; receiving, by a first device object of the plurality of device objects, a status update from a client device associated with a first user; transmitting, by the first device object, a first message comprising the status update to a second device object associated with a second user over a real-time link; and transmitting, by the first device object, a second message comprising the status update to a third device object associated with a third user over a first rate-limited link.

Abstract: Aspects of the present disclosure involve a system and a method for performing operations comprising: storing, on a distributed storage system, a plurality of real-time graph (RTG) instances that include a plurality of device objects, the plurality of device objects comprising a first device object associated with a given user; receiving, by the first device object, a friends list of the given user having a first version identifier; receiving, by the first device object, an update from a client device associated with the given user, the update comprising a friends list version identifier; determining that the first version identifier of the friends list in the first device object mismatches the friends list version identifier in the update; and synchronizing the friends list in the first device object prior to sending one or more messages that include the update to other device objects of the plurality of device objects.

Abstract: Aspects of the present disclosure involve systems and methods for performing operations comprising: storing, on a distributed storage system, one or more front-end (FE) instances and a plurality of real-time graph (RTG) instances; receiving, by a first online device object associated with a given user, an update from a first client device; generating, by the first online device object, a message that includes the update for transmission to a plurality of friends of the given user; storing, on a given one of the plurality of RTG instances, an offline device object for a first friend of the plurality of friends and a second online device object for a second friend of the plurality of friends; and transmitting, by the first online device object, the message that includes the update to the offline device object of the first friend and the second online device object of the second friend.

Abstract: Aspects of the present disclosure involve a system and a method for performing operations comprising: storing, on a distributed storage system, a front-end (FE) instance and a plurality of real-time graph (RTG) instances, each of the plurality of RTG instances includes a plurality of device objects, the FE instance being configured to communicate with a client device associated with a first user; establishing a bi-directional streaming remote procedure call (RPC) connection between the FE instance and the plurality of RTG instances; receiving, by the FE instance, a status update from the client device; determining, by the FE instance, that a first device object corresponding to the client device is stored on a first RTG instance of the plurality of RTG instances; and transmitting a first message comprising the status update from the FE instance to the first RTG instance to update the first device object.