Abstract: Aspects of the disclosure relate to generating a puddle occupancy grid including a plurality of cells. For instance, a first probability value for a puddle being located at a first location generated using sensor data from a first sensor may be received. A second probability value for a puddle being located at a second location generating using sensor data from a second sensor different from the first sensor may be received. A first cell may be identified from the plurality of cells using the first location. The first cell may also be identified using the second location. A value for the cell may be generated using the first probability value and the second probability value.

Abstract: Example embodiments relate to microlensing for real-time sensing of stray light. An example device includes an image sensor that includes a plurality of light-sensitive pixels. The device also includes a first lens positioned over a first subset of light-sensitive pixels selected from the plurality of light-sensitive pixels. Further, the device includes a controller. The controller is configured to determine a first angle of incidence of a first light signal detected by the first subset of light-sensitive pixels. The controller is also configured to, based on the first determined angle of incidence, determine an amount of stray light incident on the image sensor.

Abstract: An example method includes using a plurality of switches corresponding to a plurality of capacitors to select a first set of capacitors for charging at a first time. Charging the first set of capacitors corresponds to sampling from a first set of adjacent light detectors. The method includes using the plurality of switches to select a second set of capacitors from the plurality of capacitors for discharging at a second time. The method includes using a sampling switch to sample an output of the second set of capacitors as they discharge. The output of the second set of capacitors corresponds to the first set of adjacent light detectors. The method includes determining, based on sampling the output of the second set of capacitors, a collective intensity of light received by the first set of adjacent light detectors.

Abstract: The present disclosure relates to methods and systems that facilitate determination of a pose of a vehicle based on various combinations of map data and sensor data received from light detection and ranging (LIDAR) devices and/or camera devices. An example method includes receiving point cloud data from a (LIDAR) device and transforming the point cloud data to provide a top-down image. The method also includes comparing the top-down image to a reference image and determining, based on the comparison, a yaw error. An alternative method includes receiving camera image data from a camera and transforming the camera image data to provide a top-down image. The method also includes comparing the top-down image to a reference image and determining, based on the comparison, a yaw error.

Abstract: Example embodiments relate to low elevation side lobe antennas with fan-shaped beams. An example radar unit may include a radiating plate having a first side and a second side with an illuminator, a waveguide horn, a waveguide opening, and a radiating sleeve extending into the first side of the radiating plate. The waveguide opening is positioned on the first end of the first side and the radiating sleeve is positioned on the second end of the first side. The radar unit also includes a metallic cover coupled to the first side of the radiating plate such that the metallic cover and the radiating plate form waveguide structures.

Abstract: The technology relates to determining whether a camera is occluded. For instance, an image may be captured using a camera having red, green, and blue pixels each including a photosensor. Output values for the photosensors of each of the red pixels, green pixels, and blue pixels may be determined for the images. The output values of the green pixels may be compared to one or more of the output values of the red pixels or the output values of the blue pixels. That the camera is occluded is determined based on the comparison.

Abstract: Aspects of the disclosure provide systems and methods for confirming the identity of a passenger and changing destination of a vehicle. This may include receiving dispatching instructions to pick up a first passenger at a pickup location and to drop off the first passenger at a first destination as well as authentication information for authenticating a first client computing device of the first passenger. Once the client device is authenticated and a second passenger enters the vehicle, the vehicle is maneuvered towards the first destination. While doing so, a location of the vehicle is compared to location information received from the client computing devices. A notification is sent to a dispatching server based on the comparison and a second destination location is received in response. The vehicle is then maneuvered towards the second destination instead of the first destination.

Abstract: Methods, systems, and apparatus, including computer programs encoded on computer storage media, for processing point cloud data using dynamic voxelization. When deployed within an on-board system of a vehicle, processing the point cloud data using dynamic voxelization can be used to make autonomous driving decisions for the vehicle with enhanced accuracy, for example by combining representations of point cloud data characterizing a scene from multiple views of the scene.

Abstract: An example method includes receiving, from one or more sensors associated with an autonomous vehicle, sensor data associated with a target object in an environment of the vehicle during a first environmental condition, where at least one sensor of the sensor(s) is configurable to be associated with one of a plurality of operating field of view volumes. The method also includes based on the sensor data, determining at least one parameter associated with the target object. The method also includes determining a degradation in the parameter(s) between the sensor data and past sensor data, where the past sensor data is associated with the target object in the environment during a second environmental condition different from the first and, based on the degradation, adjusting the operating field of view volume of the at least one sensor to a different one of the operating field of view volumes.

Abstract: A light detection and ranging (LIDAR) device scans through a scanning zone while emitting light pulses and receives reflected signals corresponding to the light pulses. The LIDAR device scans the emitted light pulses through the scanning zone by reflecting the light pulses from an array of oscillating mirrors. The mirrors are operated by a set of electromagnets arranged to apply torque on the mirrors, and an orientation feedback system senses the orientations of the mirrors. Driving parameters for each mirror are determined based on information from the orientation feedback system. The driving parameters can be used to drive the mirrors in phase at an operating frequency despite variations in moments of inertia and resonant frequencies among the mirrors.

Abstract: The disclosure provides for a system that includes one or more cabin components in a vehicle and one or more computing devices of the vehicle. The one or more computing devices are configured to determine that a vehicle will have an acceleration in a first direction at a future time and an acceleration time associated with the acceleration of the vehicle, alter the one or more cabin components of the vehicle before the acceleration time, and adjust the one or more cabin components of the vehicle based on the acceleration of the vehicle over time.

Abstract: Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for generating motion detection based on optical flow. One of the methods includes obtaining a first image of a scene in an environment taken by an agent at a first time point and a second image of the scene at a second later time point. A point cloud characterizing the scene in the environment is obtained. A predicted optical flow is determined between the first image and the second image. A respective initial flow prediction for the point that represents motion of the point between the two time points is determined. A respective ego motion flow estimate for the point that represents a motion of the point induced by ego motion of the agent is determined. A respective motion prediction that indicates whether the point was static or in motion between the two time points is determined.

Abstract: This disclosure describes devices, systems, and methods that relate to obtaining image frames with variable resolutions in synchronization with a clock source. An example device may include an image sensor, a clock input, and a controller. The controller includes at least one processor and a memory. The at least one processor is operable to execute program instructions stored in the memory so as to carry out operations. The operations include receiving, by the clock input, a clock signal. The clock signal is a periodic signal defining at least one scan interval. The operations also include during the scan interval, causing the image sensor to capture a full resolution image frame. The operations yet further include during the scan interval, causing the image sensor to capture at least one reduced resolution image frame.

Abstract: An example method includes capturing, by an image capture device, a first image having a plurality of pixels. Each pixel includes a plurality of channels, and the first image is captured in accordance with first exposure parameters. The method includes determining, by a controller of the image capture device, average pixel intensities for each of the plurality of channels. The method includes determining, by the controller, a weighted average of pixel intensities using the average pixel intensities. The method includes setting, by the controller, a gain that is proportional to a ratio of a desired average pixel intensity relative to the weighted average of pixel intensities. The method includes setting, by the controller, second exposure parameters for a second image based on the gain. The method includes capturing, by the image capture device, the second image in accordance with the second exposure parameters.

Abstract: The present disclosure relates to systems and methods that facilitate a scanning light detection and ranging (LIDAR) device configured to provide an asymmetric illumination pattern. An example system includes a rotatable base configured to rotate about a first axis and a mirror assembly. The mirror assembly is configured to rotate about a second axis, which is substantially perpendicular to the first axis. The system also includes an optical cavity coupled to the rotatable base. The optical cavity includes a photodetector and a photodetector lens arranged so as to define a light-receiving axis. The optical cavity also includes a light-emitter device and a light-emitter lens arranged so as to define a light-emission axis. At least one of the light-receiving axis or the light-emission axis forms a tilt angle with respect to the first axis.

Abstract: The disclosure provides for a wiper system for cleaning a surface of a sensor housing, such as a sensor housing positioned on top of a vehicle. The wiper system includes a plurality of windows spaced around a sensor housing, and a plurality of wipers positioned around the sensor housing, each of the wipers includes a wiper blade configured to clean a corresponding window of the plurality of windows. The wiper system further includes a drive system including a moveable part coupled to a motor. The motor is configured to drive the drive system to simultaneously rotate the plurality of wipers around the sensor housing such that the plurality of wipers remove debris from the plurality of windows.

Abstract: An autonomous vehicle detects a tailgating vehicle and uses various response mechanisms. A vehicle is identified as a tailgater based on whether its characteristics meet a variable threshold. When the autonomous vehicle is traveling at slower speeds, the threshold is defined in distance. When the autonomous vehicle is traveling at faster speeds, the threshold is defined in time. The autonomous vehicle responds to the tailgater by modifying its driving behavior. In one example, the autonomous vehicle adjusts a headway buffer (defined in time) from another vehicle in front of the autonomous vehicle. In this regard, if the tailgater is T seconds too close to the autonomous vehicle, the autonomous vehicle increases the headway buffer to the vehicle in front of it by some amount relative to T.

Abstract: The technology employs a variable motion control envelope that enables an on-board computing system of a self-driving vehicle to estimate future vehicle driving behavior along an upcoming path, in order to maintain a desired amount of control during autonomous driving. Factors including intrinsic vehicle properties, extrinsic environmental influences and road friction information are evaluated. Such factors can be evaluated to derive an available acceleration model, which defines an envelope of maximum longitudinal and lateral accelerations for the vehicle. This model, which may identify dynamically varying acceleration limits that can be affected by road conditions and road configurations, may be used by the on-board control system (e.g., a planner module of the processing system) to control driving operations of the vehicle in an autonomous driving mode.

Abstract: The present disclosure relates to a target, a method, and a system for calibrating a camera. One example embodiment includes a target. The target includes a first pattern of fiducial markers. The target also includes a second pattern of fiducial markers. The first pattern of fiducial markers is a scaled version of the second pattern of fiducial markers, such that a calibration image captured of the target simulates multiple images of a single pattern captured at multiple calibration perspectives.

Abstract: Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for training a neural network configured to receive a network input and to assign a respective score to each of a plurality of locations in the network input. In one aspect, a method includes obtaining a training input and a corresponding ground truth output; processing the training input to generate a training output; computing a loss for the training input, comprising: selecting a plurality of candidate locations; setting to zero the training scores for any location in the selected candidate locations that has a ground truth score below a threshold value; for each of a plurality of pairs of locations in the selected candidate locations: computing a pair-wise loss for the pair; and combining the pair-wise losses to compute the loss for the training input; and determining an update to the current values of the parameters.