Abstract: Techniques for assisting a passenger to identify a vehicle and for assisting a vehicle to identify a passenger are discussed herein. Also discussed herein are techniques for capturing data via sensors on a vehicle or user device and for presenting such data in various formats. For example, in the context of a ride hailing service using autonomous vehicles, the techniques discussed herein can be used to identify a passenger of the autonomous vehicle at the start of a trip, and can be used to assist a passenger to identify an autonomous vehicle that has been dispatched for that particular passenger. Additionally, data captured by sensors of the vehicle and/or by sensors of a user device can be used to initiate a ride, determine a pickup location, orient a user within an environment, and/or provide visualizations or augmented reality elements to provide information and/or enrich a user experience.

Abstract: Techniques for compensating for errors in position of a vehicle are discussed herein. In some cases, a discrepancy may exist between a measured state of the vehicle and a desired state as determined by a system of the vehicle. Techniques and methods for a planning architecture of an autonomous vehicle that is able to provide maintain a smooth trajectory as the vehicle follows a planned path or route. In some cases, a planning architecture of the autonomous vehicle may compensate for differences between an estimated state and a planned path without the use of a separate system. In this example process, the planning architecture may include a mission planning system, a decision system, and a tracking system that together output a trajectory for a drive system.

Abstract: Sensors, including time-of-flight sensors, may be used to detect objects in an environment. In an example, a vehicle may include a time-of-flight sensor that images objects around the vehicle, e.g., so the vehicle can navigate relative to the objects. Sensor data generated by the time-of-flight sensor can return unreliable pixels, e.g., in the case of over-exposure or saturation. In some examples, multiple exposures captured at different exposure times can be used to determine an overall saturation value or metric representative of the sensor data. The saturation value may be used to control parameters of the sensor. For instance, the saturation value may be used to determine power control parameters for the sensor, e.g., to reduce over- and/or under-exposure.

Abstract: A vehicle includes an expandable curtain disposed in a roof of the vehicle and is configured to selectively deploy from a stowed configuration to a deployed configuration. An expandable bladder is configured to inflate at least partially during deployment of the expandable curtain. The expandable bladder includes a neck portion mechanically coupled to the expandable curtain and a head portion extending from the neck portion.

Abstract: Techniques for determining a degraded state associated with a sensor are discussed herein. For example, a sensor associated with vehicle may captured data of an environment. A portion of the data may represent a portion of the vehicle. Data associated with a region of interest can be determined based on a calibration associated with the sensor. For example, in the context of image data, image coordinates may be used to determine a region of interest, while in the context of lidar data, a beam and/or azimuth can be used to determine a region of interest. A data metric can be determined for data in the region of interest, and an action can be determined based on the data metric. For example, the action can include cleaning a sensor, scheduling maintenance, reducing a confidence associated with the data, or slowing or stopping the vehicle.

Abstract: Techniques for monitoring and predicting vehicle health are disclosed. In some examples, sensor data (e.g., audio data) may be used to create a sensor signature associated with a vehicle component. The sensor signature may be compared with one or more second sensor signatures associated with the vehicle component over the life of the vehicle component to determine changes in an operating status associated with the vehicle component. In some examples, a machine learned model may be trained to identify a vehicle component and/or and operating status of a vehicle component based on sensor data that is inputted into the machine learned model. In this way, sensor data may be input into the machine learned model and the machine learned model may output a corresponding vehicle component and/or operating status associated with the component.

Abstract: Techniques for reducing the wear on vehicles. For instance, a vehicle may include a bidirectional vehicle that operates in a first mode at which a first end of the vehicle operates as a front of the vehicle and a second mode at which a second end of the vehicle operates as the front of the vehicle. As such, the vehicle may store data representing a first distance that the vehicle has traveled while operating in the first mode and a second distance that the vehicle has traveled while operating in the second mode. The vehicle may then detect the occurrence of an event. Based on the occurrence of the event, the vehicle may select a mode for operating the vehicle using the first distance and the second distance. For example, the vehicle may select the mode that is associated with the shortest distance.

Abstract: Techniques for determining a classification probability of an object in an environment are discussed herein. Techniques may include analyzing sensor data associated with an environment from a perspective, such as a top-down perspective, using multi-channel data. From this perspective, techniques may determine channels of multi-channel input data and additional feature data. Channels corresponding to spatial features may be included in the multi-channel input data and data corresponding to non-spatial features may be included in the additional feature data. The multi-channel input data may be input to a first portion of a machine-learned (ML) model, and the additional feature data may be concatenated with intermediate output data from the first portion of the ML model, and input into a second portion of the ML model for subsequent processing and to determine the classification probabilities. Additionally, techniques may be performed on a multi-resolution voxel space representing the environment.

Abstract: Techniques for determining error models for use in simulations are discussed herein. Ground truth perception data and vehicle perception data can be determined from vehicle log data. Further, objects in the log data can be identified as relevant objects by signals output by a planner system or based on the object being located in a driving corridor. Differences between the ground truth perception data and the vehicle perception data can be determined and used to generate error models for the relevant objects. The error models can be applied to objects during simulation to increase realism and test vehicle components.

Abstract: A monocular image often does not contain enough information to determine, with certainty, the depth of an object in a scene reflected in the image. Combining image data and LIDAR data may enable determining a depth estimate of the object relative to the camera. Specifically, LIDAR points corresponding to a region of interest (“ROI”) in the image that corresponds to the object may be combined with the image data. These LIDAR points may be scored according to a monocular image model and/or a factor based on a distance between projections of the LIDAR points into the ROI and a center of the region of interest may improve the accuracy of the depth estimate. Using these scores as weights in a weighted median of the LIDAR points may improve the accuracy of the depth estimate, for example, by discerning between a detected object and an occluding object and/or background.

Abstract: Techniques for determining an output from a plurality of sensor modalities are discussed herein. Features from a radar sensor, a lidar sensor, and an image sensor may be input into respective models to determine respective intermediate outputs associated with a tracks associated with an object and associated confidence levels. The Intermediate outputs from a radar model, a lidar model, and an vision model may be input into a fused model to determine a fused confidence level and fused output associated with the track. The fused confidence level and the individual confidence levels are compared to a threshold to generate the track to transmit to a planning system or prediction system of an autonomous vehicle. Additionally, a vehicle controller can control the autonomous vehicle based on the track and/or on the confidence level(s).

Abstract: Techniques for charging a battery associated with a vehicle are discussed herein. A dense charging station for charging the battery may have lanes arranged in parallel, and each of the lanes may have sequential charging locations. A vehicle utilizing the charging station may position itself at the first available charging location, being receiving energy, and determine if a subsequent charging station becomes available in the lane, and then position itself at the subsequent charging location, once available. Multiple charging stations may be required to maintain a threshold power state for individual vehicles in a fleet of vehicles providing a service for a geographic region. A charging coordinator may determine when a battery of a vehicle does not satisfy a threshold power state and requires a recharge. Additionally, the charging coordinator may determine a candidate charging station from among multiple charging stations associated with the geographic region.

Abstract: An electrical system can include a power supply configured to provide electrical power to components at a time at which the electrical system experiences an electrical fault. The electrical system can include a first battery electrically coupled in parallel to a second battery via an electrical bus, whereby the first and second batteries can provide electrical power to a first electrical load and a second electrical load. Upon experiencing a fault, a first circuit element can electrically decouple the first battery and the second battery by opening a circuit provided by the electrical bus, thereby isolating the first battery from the second battery. Next, the battery experiencing the fault can include a second circuit element that can electrically decouple the battery experiencing the fault from a respective electrical load, while the battery isolated from the fault can continue to provide electrical power to components.

Abstract: Testing autonomous vehicle control systems in the real world can be difficult, because creating and re-creating physical scenarios for repeated testing may be impractical. In some implementations, detailed map data and data acquired through driving in a region can be used to identify similar segments of a drivable surface. Simulation scenarios used to test one of the similar segments may be used to test other of the similar segments. The driving data may also be used to generate and/or validate the simulation scenarios, e.g., by re-creating scenarios encountered while driving in a first segment in a simulation scenario for use in the second segment and comparing simulated driving behavior with the driving data.

Abstract: A vehicle may receive sensor data captured by a sensor, determine that the sensor data represents an object in the environment, and determine a collision probability associated with a collision between the vehicle and the object. Based at least in part on the collision probability, the vehicle may determine one or more mitigating actions to perform prior to, during, and/or after the collision. The mitigating action may be associated with adjusting a hydraulic fluid pressure in at least a portion of a hydraulic fluid system of the vehicle.

Abstract: Techniques associated with improving performance and realism of simulation instances associated with simulation testing of autonomous vehicles. In some cases, a simulation system may be configured to run a pre-simulation test to identify and store occlusion data to improve the performance of subsequent simulations associated with a shared scene or route.

Abstract: Performance anomalies in autonomous vehicle can be difficult to identify, and the impact of such anomalies on systems within the autonomous vehicle may be difficult to understand. In examples, systems of the autonomous vehicle are modeled as nodes in a probabilistic graphical network. Probabilities of data generated at each of the nodes is determined. The probabilities are used to determine capabilities associated with higher level functions of the autonomous vehicle.

Abstract: A method for autonomously operating a driverless vehicle along a path between a first geographic location and a destination may include receiving communication signals from the driverless vehicle. The communication signals may include sensor data from the driverless vehicle and data indicating occurrence of an event associated with the path. The communication signals may also include data indicating that a confidence level associated with the path is less than a threshold confidence level due to the event. The method may also include determining, via a teleoperations system, a level of guidance to provide the driverless vehicle based on data associated with the communication signals, and transmitting teleoperations signals to the driverless vehicle. The teleoperations signals may include guidance to operate the driverless vehicle according to the determined level of guidance, so that a vehicle controller maneuvers the driverless vehicle to avoid, travel around, or pass through the event.

Abstract: Techniques for identifying curbs are discussed herein. For instance, a vehicle may generate sensor data using one or more sensors, where the sensor data represents points associated with a driving surface and a sidewalk. The vehicle may then quantize the points into distance bins that are located laterally along the driving direction of the vehicle in order to generate spatial lines. Next, the vehicle may determine separation points for the spatial lines, where the separation points are configured to separate the points associated with the driving surface from the points associated with the sidewalk. The vehicle may then generate, using the separation points, a curve that represents the curb between the driving surface and the sidewalk. This way, the vehicle may use the curve while navigating, such as to avoid the curb and/or stop at a location that is proximate to the curb.