Abstract: Systems and methods related to redundant battery management systems (BMS) may include an on-battery BMS, and one or more off-battery, off-vehicle, and/or remote BMS. The one or more remote BMS may receive data associated with parameters of a battery, and may process the received data using various models, algorithms, or estimators to determine a battery state. The battery state determined by the remote BMS may then be used to corroborate or support a battery state determined by the on-battery BMS, thereby ensuring safe, reliable, and efficient operations of systems, machines, or devices utilizing batteries as power sources.

Abstract: Shape memory actuators may be used in unmanned aerial vehicles to control various components. For example, shape memory actuators may adjust cant angles of motors, propellers, and other propulsion mechanisms. In addition, shape memory actuators may adjust positions or orientations of various other components of unmanned aerial vehicles, including wings, control surfaces, motor arms, frame sections, payload doors, and landing gears. The shape memory actuators may be formed of various shape memory materials, may be one-way or two-way shape memory actuators, and may change their configurations responsive to heat and/or magnetic fields.

Abstract: An aerial vehicle outfitted with one or more cameras captures a sequence of images of a rotating propeller. The images are processed according to one or more techniques to recognize a position of the propeller, or an angle of orientation of the propeller, in each of the images. The positions or angles of the rotating propeller are used to calculate a rotational speed of the propeller, or a direction of rotation of the propeller. The cameras used to capture the images are also available for use in other applications such as navigation, monitoring or collision avoidance. The propeller may include one or more stripes, colors, characters, symbols or other markings to enhance its visibility and facilitate the determination of positions or angles of the propeller within images.

Abstract: A convolutional neural network may be trained to inspect subjects such as carbon fiber propellers for surface flaws or other damage. The convolutional neural network may be trained using images of damaged and undamaged subjects. The damaged subjects may be damaged authentically during operation or artificially by manual or automated means. Additionally, images of undamaged subjects may be synthetically altered to depict damages, and such images may be used to train the convolutional neural network. Images of damaged and undamaged subjects may be captured for training or inspection purposes by an imaging system having cameras aligned substantially perpendicular to subjects and planar light sources aligned to project light upon the subjects in a manner that minimizes shadows and specular reflections. Once the classifier is trained, patches of an image of a subject may be provided to the classifier, which may predict whether such patches depict damage to the subject.

Abstract: Example variable pitch propulsion mechanisms may include propeller hubs having variable pitch propeller blades. The propeller hubs may include shape memory actuators operatively connected to the propeller blades to change their pitch. In addition, the propeller hubs may include linkage arms to transfer movement from the shape memory actuators to the propeller blades. Further, the propeller hubs may include geared connections to transfer movement from the shape memory actuators to the propeller blades. Moreover, the propeller hubs may include latch mechanisms, similar to retractable pen mechanisms, to transfer movement from the shape memory actuators to the propeller blades and decouple movement of the propeller blades from alignment of the propeller blades.

Abstract: A convolutional neural network may be trained to inspect subjects such as carbon fiber propellers for surface flaws or other damage. The convolutional neural network may be trained using images of damaged and undamaged subjects. The damaged subjects may be damaged authentically during operation or artificially by manual or automated means. Additionally, images of undamaged subjects may be synthetically altered to depict damages, and such images may be used to train the convolutional neural network. Images of damaged and undamaged subjects may be captured for training or inspection purposes by an imaging system having cameras aligned substantially perpendicular to subjects and planar light sources aligned to project light upon the subjects in a manner that minimizes shadows and specular reflections. Once the classifier is trained, patches of an image of a subject may be provided to the classifier, which may predict whether such patches depict damage to the subject.