Abstract: An unmanned aerial vehicle with deployable components (UAVDC) may comprise a foldable propeller blade with a locking mechanism. Foldable propeller blades may have a stowed configuration and a deployed configuration relative to the UAVDC, and the foldable propeller blades may pivot about a hinge to move between configurations. In the deployed configuration, the foldable propeller may experience forward folding forces acting upon it. The locking mechanism may lock the foldable propeller blade in the deployed configuration. The locking mechanism may keep the foldable propeller locked into place to prevent forward folding tendency.

Abstract: Aspects of the subject technology relate to systems and methods for controlling a hydraulic fracturing job.

Abstract: A timing adjustment assembly for a compound archery bow is provided. The timing adjustment assembly is coupled with a pulley assembly and power cable of the archery bow. The timing adjustment assembly includes an adjustable component and a stationary component. The stationary component is attached to either the pulley assembly, the power cable, or the riser of the archery bow. The adjustable component is configured to move in a linear direction between multiple positional arrangements relative to the stationary component. The movement of the adjustable component changes the effective length of the power cable by moving the termination anchor point of the power cable or deflecting the power cable. The timing adjustment assembly adjusts the effective length of the power cable and relative synchronization of the pulley assembly with the archery bow in a brace position through the change in termination anchor point or deflection.

Abstract: Techniques are described herein that are capable of tracking an eye of a user using multiple lasers. Light from the lasers is scanned across respective partially overlapping portions of a region that includes an eye of a user during respective time periods. Portion(s) of the light that are reflected from the eye are detected by respective photodetector(s). In an example implementation, a signal corresponding to the detected portion(s) is provided in a pixel of a frame buffer based at least in part on a current angle of a mirror used to scan the light across the region. In a second implementation, digital state(s) are provided based at least in part on difference(s) between a reference signal and signal(s) corresponding to the detected portion(s), and a time value indicating a time at which a glint is detected by a photodetector is provided when a digital state triggers an interrupt handler.

Abstract: Techniques are described herein that are capable of tracking an eye of a user using multiple lasers. Light from the lasers is scanned across respective partially overlapping portions of a region that includes an eye of a user during respective time periods. Portion(s) of the light that are reflected from the eye are detected by respective photodetector(s). In an example implementation, a signal corresponding to the detected portion(s) is provided in a pixel of a frame buffer based at least in part on a current angle of a mirror used to scan the light across the region. In a second implementation, digital state(s) are provided based at least in part on difference(s) between a reference signal and signal(s) corresponding to the detected portion(s), and a time value indicating a time at which a glint is detected by a photodetector is provided when a digital state triggers an interrupt handler.

Abstract: A MEMS scanning device (“Device”) includes at least (1) laser projector(s) controlled by a laser drive to project a laser beam, (2) MEMS scanning mirror(s) controlled by a MEMS drive to scan the laser beam to generate a raster scan, (3) a display configured to receive the raster scan, (4) a thermometer configured to detect a current temperature, (5) a display observing camera configured to capture an image of a predetermined area of the display, and (6) a computer-readable media that stores temperature model(s), each of which is custom-built using machine learning. The device uses the display observing camera to capture image(s) of predetermined pattern(s), which are then used to extract feature(s). The extracted feature(s) are compared with ideal feature(s) to identify a discrepancy. When the identified discrepancy is greater than a threshold, the temperature model(s) are updated accordingly.

Abstract: A MEMS scanning device (“Device”) includes at least (1) laser projector(s) controlled by a laser drive to project a laser beam, (2) MEMS scanning mirror(s) controlled by a MEMS drive to scan the laser beam to generate a raster scan, (3) a display configured to receive the raster scan, (4) a thermometer configured to detect a current temperature, (5) a display observing camera configured to capture an image of a predetermined area of the display, and (6) a computer-readable media that stores temperature model(s), each of which is custom-built using machine learning. The device uses the display observing camera to capture image(s) of predetermined pattern(s), which are then used to extract feature(s). The extracted feature(s) are compared with ideal feature(s) to identify a discrepancy. When the identified discrepancy is greater than a threshold, the temperature model(s) are updated accordingly.

Abstract: A MEMS scanning device (“Device”) includes at least (1) laser projector(s) controlled by a laser drive to project a laser beam, (2) MEMS scanning mirror(s) controlled by a MEMS drive to scan the laser beam to generate a raster scan, (3) a display configured to receive the raster scan, (4) a thermometer configured to detect a current temperature, (5) a display observing camera configured to capture an image of a predetermined area of the display, and (6) a computer-readable media that stores temperature model(s), each of which is custom-built using machine learning. The device uses the display observing camera to capture image(s) of predetermined pattern(s), which are then used to extract feature(s). The extracted feature(s) are compared with ideal feature(s) to identify a discrepancy. When the identified discrepancy is greater than a threshold, the temperature model(s) are updated accordingly.

Abstract: Techniques are provided to re-arrange the placement of a photodiode within an illumination system to achieve improved characteristics and reduced form factor. An illumination system includes a laser assembly, a MEMS mirror system, a beam combiner, and a photodiode. The laser assembly includes RGB lasers, and the MEMS mirror system redirects laser light produced by the RGB lasers to illuminate pixels in an image frame. The beam combiner combines the laser light. The photodiode is provided to determine a power output of the laser assembly by receiving and measuring some of the laser light. The photodiode may be beneficially positioned before or after collimating optics and/or the beam combiner.

Abstract: A MEMS scanning device (“Device”) includes at least (1) laser projector(s) controlled by a laser drive to project a laser beam, (2) MEMS scanning mirror(s) controlled by a MEMS drive to scan the laser beam to generate a raster scan, (3) a display configured to receive the raster scan, (4) a thermometer configured to detect a current temperature, (5) a display observing camera configured to capture an image of a predetermined area of the display, and (6) a computer-readable media that stores temperature model(s), each of which is custom-built using machine learning. The device uses the display observing camera to capture image(s) of predetermined pattern(s), which are then used to extract feature(s). The extracted feature(s) are compared with ideal feature(s) to identify a discrepancy. When the identified discrepancy is greater than a threshold, the temperature model(s) are updated accordingly.

Abstract: A display device includes a light source, a support structure, and a scanning mirror system. The scanning mirror system includes a mirror, a first anchor located at a first lateral side of the scanning mirror system, a second anchor located at a second lateral side of the scanning mirror system, and a flexure. A first portion of the flexure extends from the first anchor toward a first longitudinal end and turns to meet a first end of the mirror. A second portion extends from the second anchor toward a second longitudinal end and turns to meet to a second end of the mirror opposite the first end. An actuator system is configured to actuate the flexure to thereby vary a scan angle of the mirror.

Abstract: A MEMS scanner may include a first flexible arm extending substantially in a forward direction and a base connected to a proximal end of the first flexible arm, the base being thicker than the first flexible arm in a vertical direction. The MEMS scanner may further include a second flexible arm connected to a distal end of the first flexible arm, the second flexible arm extending substantially in a reverse direction. The MEMS scanner may further include a mirror coupled to a distal end of the second flexible arm. In one implementation, the MEMS scanner may be a non-resonant scanner.

Abstract: Techniques are provided to reduce the form factor of laser-based systems by multi-purposing a photodiode used to help control the output of a laser. A reflective photodiode comprises a light receiving surface and a reflective coating. The light receiving surface is configured to absorb some incident light and to convert it into electrical current. The reflective coating is disposed on the light receiving surface and is configured to reflect some of the incident light away from the light receiving surface. The reflective coating also permits some of the incoming light to pass therethrough for absorption.

Abstract: A MEMS scanner may include a first flexible arm extending substantially in a forward direction and a base connected to a proximal end of the first flexible arm, the base being thicker than the first flexible arm in a vertical direction. The MEMS scanner may further include a second flexible arm connected to a distal end of the first flexible arm, the second flexible arm extending substantially in a reverse direction. The MEMS scanner may further include a mirror coupled to a distal end of the second flexible arm. In one implementation, the MEMS scanner may be a non-resonant scanner.

Abstract: A MEMS scanning device (“Device”) includes at least (1) laser projector(s) controlled by a laser drive to project a laser beam, (2) MEMS scanning mirror(s) controlled by a MEMS drive to scan the laser beam to generate a raster scan, (3) a display configured to receive the raster scan, (4) a thermometer configured to detect a current temperature, (5) a display observing camera configured to capture an image of a predetermined area of the display, and (6) a computer-readable media that stores temperature model(s), each of which is custom-built using machine learning. The device uses the display observing camera to capture image(s) of predetermined pattern(s), which are then used to extract feature(s). The extracted feature(s) are compared with ideal feature(s) to identify a discrepancy. When the identified discrepancy is greater than a threshold, the temperature model(s) are updated accordingly.

Abstract: Variable attenuation of an illumination source is provided by an aperture (i.e., an opening in a structure through which light passes) that is sandwiched between two tunable lenses that are configured to apply varying amounts of optical power. A controller operates the first tunable lens to apply optical power to the light to be divergent at the aperture structure so that a portion of the light is clipped. Varying the applied optical power at the first tunable lens can increase or decrease divergence at the aperture structure to thereby increase or decrease clipping and the attenuation of the light. The controller operates the second tunable lens to compensate for changes in light state at the first tunable lens by applying opposite optical power so that collimated light from the illumination source which enters the first tunable lens may exit the second tunable lens in the same collimated state.

Abstract: Examples are disclosed that relate to scanning display systems. One example provides a display device comprising a controller, a light source, and a scanning mirror system. The scanning mirror system comprises a scanning mirror configured to scan light from the light source in at least one direction at a resonant frequency of the scanning mirror, and an electromechanical actuator system coupled with the scanning mirror and being controllable by the controller to adjust the resonant frequency of the scanning mirror.

Abstract: Techniques are described herein that are capable of tracking an eye of a user using multiple lasers. Light from the lasers is scanned across respective partially overlapping portions of a region that includes an eye of a user during respective time periods. Portion(s) of the light that are reflected from the eye are detected by respective photodetector(s). In an example implementation, a signal corresponding to the detected portion(s) is provided in a pixel of a frame buffer based at least in part on a current angle of a mirror used to scan the light across the region. In a second implementation, digital state(s) are provided based at least in part on difference(s) between a reference signal and signal(s) corresponding to the detected portion(s), and a time value indicating a time at which a glint is detected by a photodetector is provided when a digital state triggers an interrupt handler.

Abstract: Aspects of the present disclosure provide methods and systems for training a neural network. Embodiments include determining a symmetric key that indicates a scrambled ordering of data components. Embodiments include receiving a plurality of training data instances. Embodiments include training a neural network by, for each respective training data instance of the plurality of training data instances: identifying a training input and a training output of the respective training data instance; identifying a plurality of training input components of the training input; providing the plurality of training input components to an input layer of the neural network based on the scrambled ordering indicated by the symmetric key; receiving an output from the neural network in response to the plurality of training input components; and determining whether to modify one or more parameters of the neural network based on the output and the training output of the respective training data instance.