Abstract: Methods, systems, and apparatus, including computer programs encoded on computer storage media, for using a surfel map to generate a prediction for a state of an environment. One of the methods includes obtaining surfel data comprising a plurality of surfels, wherein each surfel corresponds to a respective different location in an environment, and each surfel has associated data that comprises an uncertainty measure; obtaining sensor data for one or more locations in the environment, the sensor data having been captured by one or more sensors of a first vehicle; determining one or more particular surfels corresponding to respective locations of the obtained sensor data; and combining the surfel data and the sensor data to generate a respective object prediction for each of the one or more locations of the obtained sensor data.

Abstract: Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for generating object interaction predictions using a neural network. One of the methods includes obtaining a sensor input derived from data generated by one or more sensors that characterizes a scene. The sensor input is provided to an object interaction neural network. The object interaction neural network is configured to process the sensor input to generate a plurality of object interaction outputs. Each respective object interaction output includes main object information and interacting object information. The respective object interaction outputs corresponding to the plurality of regions in the sensor input are received as output of the object interaction neural network.

Abstract: Systems and methods described herein relate to LIDAR systems and their operation. An example method includes partitioning a plurality of light-emitter devices into a plurality of groups. Each light-emitter device is associated with a given group of the plurality of groups. The method also includes selecting a group from the plurality of groups according to a predetermined group order and selecting one or more light-emitter devices from the plurality of light-emitter devices of the selected group according to a firing order. The method yet further includes, at a predetermined shot dither time, causing the selected light-emitter device to emit at least one light pulse. The predetermined shot dither time is based on a shot dither schedule. The method may additionally include repeating the method to provide a complete scan in which each light-emitter device of the plurality of light-emitter devices has emitted at least one light pulse.

Abstract: Aspects of the disclosure provide for reducing inconvenience to other road users caused by stopped autonomous vehicles. As an example, a vehicle having an autonomous driving mode may be stopped at a first location. While the vehicle is stopped, sensor data is received from a perception system of the vehicle. The sensor data may identify a road user. Using the sensor data, a value indicative of a level of inconvenience to the road user caused by stopping the vehicle at the first location may be determined. The vehicle is controlled in the autonomous driving mode to cause the vehicle to move from the first location and in order to reduce the value.

Abstract: The technology relates to detecting and responding to processions. For instance, sensor data identifying two or more objects in an environment of a vehicle may be received. The two or more objects may be determined to be disobeying a predetermined rule in a same way. Based on the determination that the two or more objects are disobeying a predetermined rule, that the two or more objects are involved in a procession may be determined. The vehicle may then be controlled autonomously in order to respond to the procession based on the determination that the two or more objects are involved in a procession.

Abstract: Example embodiments relate to using remote assistance to maneuver an autonomous vehicle to a location. A computing device used by a remote operator may receive a request for assistance from a vehicle that indicates the vehicle is stopped at a first location with one or more navigation options for enabling the vehicle to navigate from the first location to a second location. At least one navigation option includes a maneuver technique that requires operator approval prior to execution. The computing device may then display a graphical user interface (GUI) that conveys the one or more navigation options. Based on detecting a selection of a particular navigation option, the computing device may transmit instructions to the vehicle to perform the particular navigation option. The vehicle may configured to navigate from the first location to the second location by performing the particular navigation option while monitoring for changes in the environment.

Abstract: The present disclosure relates to systems, methods, and vehicles that facilitate a light detection and ranging (LIDAR or lidar) system that may take advantage of “dead angles” where the lidar system is oriented toward support structure or another “uninteresting” feature. In such scenarios, light pulses may be redirected toward more-interesting features in the environment. An example system includes a lidar system configured to emit light pulses into an environment of the system so as to provide information indicative of objects within a default field of view. The system also includes a reflective surface optically coupled to the lidar system. The reflective surface is configured to reflect at least a portion of the emitted light pulses so as to provide an extended field of view. The lidar system is further configured to provide information indicative of objects within the extended field of view.

Abstract: The present disclosure relates to limitation of noise on light detectors using an aperture. One example embodiment includes a system. The system includes a lens disposed relative to a scene and configured to focus light from the scene onto a focal plane. The system also includes an aperture defined within an opaque material disposed at the focal plane of the lens. The aperture has a cross-sectional area. In addition, the system includes an array of light detectors disposed on a side of the focal plane opposite the lens and configured to intercept and detect diverging light focused by the lens and transmitted through the aperture. A cross-sectional area of the array of light detectors that intercepts the diverging light is greater than the cross-sectional area of the aperture.

Abstract: A route for a trip to a destination is generated using map information. A set of no-go roadway segments, where the vehicle is not able to drive in an autonomous mode, relevant to the route from the plurality of no-go roadway segments is identified from the map information. A local region around a current location of the vehicle is determined. A local map region including roadway segments of the map information that correspond to locations within the local region is determined. The set of the plurality of no-go roadway segments is filtered from the roadway segments of the local map region. A cost value is assigned to each roadway segment of the filtered roadway segments of the local map region. Any assigned cost values are used to determining a plan for maneuvering the vehicle for a predetermined period into the future. The vehicle is maneuvered according to the plan.

Abstract: Example embodiments relate to LIDAR systems with multi-faceted mirrors. An example embodiment includes a LIDAR system. The system includes a multi-faceted mirror that includes a plurality of reflective facets, which rotates about a first rotational axis. The system also includes a light emitter configured to emit a light signal toward one or more regions of a scene. Further, the system includes a light detector configured to detect a reflected light signal. In addition, the system includes an optical window positioned between the multi-faceted mirror and the one or more regions of the scene such that light reflected from one or more of the reflective facets is transmitted through the optical window. The optical window is positioned such that the optical window is non-perpendicular to the direction toward which the light emitted along the optical axis is directed for all angles of the multi-faceted mirror.

Abstract: Rotatable mirror assemblies and light detection and ranging systems containing rotatable mirror assemblies are described herein. An example rotatable mirror assembly may include (1) a housing having a top end, a bottom end, and a longitudinal axis intersecting the top and bottom ends, and (2) a set of reflective surfaces, where each reflective surface in the set is coupled to the top end of the housing and the bottom end of the housing such that each reflective surface possesses limited freedom of movement with respect to the housing.

Abstract: Example embodiments relate to substrate integrated waveguide (SIW) transitions. An example SIW may include a dielectric substrate having a top surface and a bottom surface and a first metallic layer portion coupled to the top surface of the dielectric substrate that includes a single-ended termination, an impedance transformer, and a metallic rectangular patch located within an open portion in the first metallic layer portion such that the open portion forms a non-conductive loop around the metallic rectangular patch. The SIW also includes a second metallic layer portion coupled to the bottom surface of the dielectric substrate and metallic via-holes electrically coupling the first metallic layer to the second metallic layer. The SIW may be implemented in a radar unit to couple antennas to a printed circuit board (PCB). In some examples, the SIW may be implemented with only a non-conductive opening that lacks the metallic rectangular patch.

Abstract: The technology relates to a system for clearing a sensor cover. The system may comprise a wiper comprising a wiper support, a wiper blade, and a sensor cover. The wiper blade may be configured to clear the sensor cover of debris, and the sensor cover may be configured to house one or more sensors. A wiper motor may rotate the wiper and a sensor motor may rotate the sensor cover. The system wiper blade may comprise a first edge attached to the wiper support and a second edge which may be configured to be in contact with the sensor cover. The wiper blade may extend in a corkscrew shape around the wiper support. The wiper motor may be configured to rotate the wiper in a first direction and the sensor motor may be configured to rotate the sensor cover in a second direction opposite the first direction.

Abstract: Examples relating to vehicle velocity calculation using radar technology are described. An example method performed by a computing system may involve, while a vehicle is moving on a road, receiving, from two or more radar sensors mounted at different locations on the vehicle, radar data representative of an environment of the vehicle. The method may involve, based on the data, detecting at least one scatterer in the environment. The method may involve making a determination of a likelihood that the at least one scatterer is stationary with respect to the vehicle. The method may involve, based on the determination being that the likelihood is at least equal to a predefined confidence threshold, calculating a velocity of the vehicle based on the data from the sensors. The calculated velocity may include an angular and linear velocity. Further, the method may involve controlling the vehicle based on the calculated velocity.

Abstract: A method includes determining a target specification map that is associated with a task and that indicates, for each respective region of a plurality of regions around a vehicle equipped with a sensor, a target value of a parameter of the sensor. The method also includes determining a capability specification map that indicates, for each respective region, an attained value of the parameter that the sensor is configured to provide. The method additionally includes comparing the capability specification map to the target specification map to determine, for each respective region, a disparity between the target value and the attained value. The method further includes, based on the comparing, identifying one or more of: a first subset of the plurality of regions where the target value exceeds the attained value or a second subset of the plurality of regions where the attained value meets or exceeds the target value.

Abstract: A neural network system for identifying positions of objects in an input image can include an object detector neural network, a memory interface subsystem, and an external memory. The object detector neural network is configured to, at each time step of multiple successive time steps, (i) receive a first neural network input that represents the input image and a second neural network input that identifies a first set of positions of the input image that have each been classified as showing a respective object of the set of objects, and (ii) process the first and second inputs to generate a set of output scores that each represents a respective likelihood that an object that is not one of the objects shown at any of the positions in the first set of positions is shown at a respective position of the input image that corresponds to the output score.

Abstract: Example embodiments relate to beam homogenization for occlusion avoidance. One embodiment includes a light detection and ranging (LIDAR) device. The LIDAR device includes a transmitter and a receiver. The transmitter includes a light emitter. The light emitter emits light that diverges along a fast-axis and a slow-axis. The transmitter also includes a fast-axis collimation (FAC) lens optically coupled to the light emitter. The FAC lens is configured to receive light emitted by the light emitter and reduce a divergence of the received light along the fast-axis of the light emitter to provide reduced-divergence light. The transmitter further includes a transmit lens optically coupled to the FAC lens. The transmit lens is configured to receive the reduced-divergence light from the FAC lens and provide transmit light. The FAC lens is positioned relative to the light emitter such that the reduced-divergence light is expanded at the transmit lens.

Abstract: The disclosure provides for a system. The system may include a rotational control system configured to rotate a seat of a vehicle, and one or more computing devices. The one or more computing devices may have one or more processors that are configured to determine that an impact is imminent at a location on the vehicle along a collision axis. A most favorable orientation may be determined by the one or more processors based on the determined location and collision axis. Using the rotational control system, the one or more processors may rotate the seat of the vehicle to the most favorable orientation in order to reduce risks of serious injury to a passenger in the seat caused by the imminent impact. A translational control system may be used by the one or more processors to translate the seat of the vehicle to a position relative to the determined location.

Abstract: Systems and methods are described that relate to a scanning laser system configured to emit laser light and an interlock circuit communicatively coupled to the scanning laser system. The interlock circuit may carry out certain operations. The operations include, as the scanning laser system emits laser light into one or more regions of an environment around the scanning laser system, determining a respective predicted dosage amount for each region based on the emitted laser light. The operations further include detecting an interlock condition. The interlock condition includes a predicted dosage amount for at least one region being greater than a threshold dose. In response to detecting the interlock condition, the operations include controlling the scanning laser system to reduce a subsequent dosage amount in the at least one region.

Abstract: One example system comprises a light source configured to emit light. The system also comprises a waveguide configured to guide the emitted light from a first end of the waveguide toward a second end of the waveguide. The waveguide has an output surface between the first end and the second end. The system also comprises a plurality of mirrors including a first mirror and a second mirror. The first mirror reflects a first portion of the light toward the output surface. The second mirror reflects a second portion of the light toward the output surface. The first portion propagates out of the output surface toward a scene as a first transmitted light beam. The second portion propagates out of the output surface toward the scene as a second transmitted light beam.