Wing Aviation LLC

A technique for processing delivery aborts by a UAV delivery service includes: acquiring a delivery zone (DZ) image of a delivery destination including a DZ for a package being delivered to the delivery destination by a UAV; determining to abort a delivery mission for the package based upon the DZ image; converting the DZ image to a vector embedding; performing a similarity search on a vector database using the vector embedding, the vector database storing reference images of other delivery destinations indexed to reference vector embeddings and outcome attributes describing delivery outcomes associated with the reference images, and wherein the similarity search identifies a subset of the reference images deemed to have a threshold similarity to the DZ image; and prompting a vision language model with the DZ image and the subset of the reference images to provide an abort explanation or to determine a delivery disposition for the DZ.

A technique for provisioning UAVs to operate in a specific geographical region includes acquiring reference aerial images of the specific geographical region from a secondary source. The reference aerial images have at least one style characteristic that is distinct from that of operational aerial images acquired by the UAVs during operation. The reference aerial images are transformed into training images that adopt a style of the operational aerial images using a neural style transfer. A neural network that facilitates vision-based navigation of the UAVs is trained using the training images. The neural network is trained to operate in the specific geographical region prior to operation in the specific geographical region.

A technique for informing navigation of a UAV includes storing an asset map of a ground area, wherein the asset map includes a reference aerial image of the ground area annotated with labels describing reference objects depicted in the reference aerial image; acquiring a current aerial image of the ground area with an onboard camera system of the UAV while the UAV is flying above the ground area; mapping correspondences between the reference aerial image and the current aerial image using a homography estimating tool executing onboard the UAV; analyzing the current aerial image with an object detection model to detect a first object positioned at the ground area; and validating or informing a detection of the first object by the object detection model based on the mapping of the correspondences.

A method includes receiving an input specifying a starting location and a destination location for an aerial vehicle. The method additionally includes determining, based on the starting location and the destination location, an aerial path for the aerial vehicle to follow from the starting location to the destination location. The method also includes determining, based on the aerial path, a property of aerial image data, where the aerial image data is obtainable using the aerial vehicle while traversing the aerial path, and where the aerial image data represents an environment along the aerial path. The method further includes determining, based on the property, a path score associated with the aerial path, and outputting the aerial path based on the path score.

In some embodiments, a computer-implemented method of communicating sensor data between an autonomous vehicle computing system and a remote computing system is provided. The autonomous vehicle computing system gathers sensor data generated by at least one sensor of the autonomous vehicle computing system. The autonomous vehicle computing system determines an amount of available bandwidth between the autonomous vehicle computing system and the remote computing system. The autonomous vehicle computing system retrieves a local portion of a split machine learning model corresponding to the amount of available bandwidth from a model data store. The autonomous vehicle computing system processes the sensor data using the local portion to produce an intermediate result. The autonomous vehicle computing system transmits the intermediate result to the remote computing system for processing of the intermediate result using a remote portion of the split machine learning model.

An unmanned aerial vehicle (UAV) includes a fuselage including a front, a rear, a top, a bottom, and an open cargo bay with a lower access opening in the bottom of the fuselage. An aerodynamic device is positioned on a surface of the bottom of the fuselage in front of the open cargo bay. A coupler is configured to secure a payload at the top of the open cargo bay.

A method includes determining, by an unmanned aerial vehicle (UAV), a position of an autoloader device for the UAV; based on the determined position of the autoloader device, causing the UAV to follow a descent trajectory in which the UAV moves from a starting position to a first nudged position in order to deploy a tethered pickup component of the UAV to a payout position on an approach side of the autoloader device; deploying the tethered pickup component of the UAV to the payout position; causing the UAV to follow a side-step trajectory in which the UAV moves laterally to a second nudged position in order to cause the tethered pickup component of the UAV to engage the autoloader device; and retracting the tethered pickup component of the UAV to pick up a payload from the autoloader device.

A technique for maintaining a terrain model includes: acquiring aerial images of a scene at an area of interest (AOI), wherein the aerial images are acquired by a UAV during a flight mission of the UAV that passes over the AOI; training a ML model onboard the UAV with one or more of the aerial images, wherein the ML model comprises a neural network, which after the training, encodes a volumetric representation of the scene; determining whether the terrain model of the AOI is deemed out-of-date based upon whether the training results in greater than a threshold change in the ML model; and uploading image data acquired by the UAV during the flight mission to a backend data system in response to determining that the terrain model is deemed out-of-date, wherein the image data includes, or is derived from, at least a portion of the aerial images.

A technique for localizing utility lines includes capturing aerial images of a ground area below a UAV; recording a position of the UAV when capturing the aerial images; detecting a presence of an object in the aerial images suspected to be a utility line; identifying two offset pixel points in each of the aerial images that coincide with the object in each aerial image; converting the two offset pixel points in each of the aerial images to a world frame; defining a plurality of geometric planes in the world frame each corresponding to one the aerial images, wherein the each of the geometric planes is defined by the position of the UAV when capturing a corresponding to one of the aerial images and the two offset pixel points in the world frame corresponding to the one of the aerial images; and determining an intersection approximation of the geometric planes.

A joiner includes a first and second portions configured to mount to respective first and second structural members. Each of the first and second portions includes a body with a receiving surface to receive the respective structural member. The first portion includes a group of first coupling structures disposed on the first body and the second portion includes a group of second coupling structures disposed on the second body. The first coupling structures are configured to mate with the second coupling structures such that engagement surfaces of the first coupling structures engage engagement surfaces of the second coupling structures so as to transfer loads between the second portion and first portion and to limit relative movement between the first portion and second portion with respect to the first direction and with respect to a vertical direction.