Abstract: An autonomous vehicle (AV) detects an animal in its vicinity, classifies the detected animal, and selects an action to take based on the classification. The action may be selected to deter the animal away from a roadway along which the AV is traveling, or for the AV to avoid a collision or other negative encounter with the animal. The AV determines whether to proceed along the roadway, e.g., if the action has caused the animal to leave the roadway. The animal classification may be used to determine whether or not to drop off or pick up a user at a particular location near an animal.

Abstract: A method of managing a condition of a shared ride provided using an autonomous vehicle (AV) is described and includes determining individual preferences for the condition expressed by passengers participating in the shared ride; determining a consensus preference for the condition based on the individual preferences; and applying the consensus preference for the condition for at least a portion of the shared ride, wherein the applying the consensus preference for the condition comprises managing settings of the AV, including settings of a cabin of the AV.

Abstract: A phase modulated continuous wave (PMCW) multiple input multiple output (MIMO) radar system is described herein. The radar system is configured to compute range, velocity, and direction of arrival angle of objects relative to the radar system. The radar system includes several transmitting antennas and several receiving antennas, where selected transmitting antennas simultaneously transmit radar signals based on the same modulation signal. Per transmitting antenna, the transmissions are modulated with respective phase offsets on a per pulse repetition interval (PRI) basis. Hence, a coupling between phase shifts over PRI and transmitter positions is established. Effectively, then, each transmitting antenna is labeled with a velocity offset that corresponds to the phase rate of change assigned to the transmitting antenna. This approach is referred to herein as velocity-labeled multiplexing (VLM).

Abstract: A method for detecting AV location data includes receiving a data set from a data store storing data in connection with an autonomous vehicle (AV) fleet, the received data set comprising a plurality of table records; filtering the received data set by comparing data values of the plurality of table records with ranges of latitude values and longitude values and removing table records comprising data values outside both the range of latitude values and the range of longitude values; post-processing the filtered data set to identify tables that include a first table record comprising a data value within the range of latitude values and a corresponding second table record comprising a data value within the range of longitude values; and storing information regarding the identified tables in a data map, wherein the identified tables are tagged to indicate that the identified tables include location information.

Abstract: Systems and methods for automated surveillance of a fleet of vehicles in a parking area using vehicle sensors on various fleet vehicles. In particular, the vehicles in a parking area can be observed and monitored using sensors on a subset of the vehicles. The field of view of each vehicle and overall map coverage of a parking area can be used to determine which vehicles to use for surveillance and/or which sensors on various vehicles to use for surveillance. Thus, surveillance can be optimized using fleet knowledge of vehicle locations within a facility and the scope of coverage of various vehicle sensors to determine where there is overlap in coverage. Appropriate sensors on certain vehicles can be selectively activated to maintain secure coverage without burning resources unnecessarily.

Abstract: A computer-implemented method, comprising: dividing a first binary image into a plurality of variable-sized chunks, wherein the first binary image is an aggregate of a plurality of files, and wherein the dividing does not depend on file boundaries; and computing hashes of the variable-sized chunks, and storing the hashes in a content addressable storage (CAS) with the hashes as keys.

Abstract: A light inspection system positions an autonomous vehicle (AV) in a field of view of a camera such that the camera captures an image of a light of the AV. The light inspection system instructs a camera to capture an image of the light while the light is switched on. The light inspection system receives the captured image, determines a luminance of the light based on the image, and determines to service the light in response to the luminance of the light being below a threshold luminance.

Abstract: Technologies disclosed herein facilitate identification of a trip route from an origin of a user to a desired destination of the user, wherein the trip route includes multiple transportation modalities, at least one of which is an autonomous vehicle (AV), and dispatching of the AV to a pickup location for the user in connection with providing transportation to the user along a portion of the identified trip route.

Abstract: A method is described and includes subsequent to an autonomous vehicle becoming immobilized, initiating a local assistance request; subsequent to the initiating, receiving local assistance input from a passenger of the autonomous vehicle; and using the local assistance input to determine an action to be taken by the autonomous vehicle to mobilize the autonomous vehicle.

Abstract: The subject technology is related to autonomous vehicles (AV) and, in particular, to an environmental and electromagnetic seal for autonomous vehicle control systems. The AV includes an autonomous driver system computer (ADSC) fixed to an interior surface of the AV and configured to control the electronic drivetrain, the ADSC including at least one port for connecting with an AV and an elastic conductive cover for selectively sealing the at least port one. The elastic conductive cover comprises a hooking member with an opening that is rotatably connected to the ADSC, cannot be independently removed from the ADSC, and includes an inner surface having at least one extended member with a sealing member that is configured to seal a port of the at least one port.

Abstract: A control loop for scaling cloud-based agents in a cluster includes a controller and a feedback signal. The controller may be a proportional-integral controller, or a proportional-integral-derivative controller. The controller may calculate a control variable based on a target number of idle agents for the cluster. The controller receives a target number of idle agents, as well as a current number of idle agents in a cluster. The controller computes the control variable based on the error between the target and current numbers of idle agents, and an integral of the error over time.

Abstract: The disclosed technology provides solutions for generating accurate virtual representations of real-world environments. A process of the disclosed technology can include steps for: processing image sensor data to identify one or more images of road paint; identifying a geographic location that is associated with the one or more images of road paint; and generating, within a simulated environment corresponding to the geographic location, at least one road paint mesh object that is based on the one or more images of road paint.

Abstract: The navigation of an AV in an environment may be planned based on a model of a restricted traffic zone in the environment. Information of the environment (e.g., a vector map, information of one or more objects, a temporal sequence of semantic grids, a query grid, etc.) may be input into a neural network. The neural network may include a CNN and a GNN. The map and tracks may be input into the GNN. The temporal sequence of semantic grids or query grid may be input into the CNN. The neural network may output edges of the restricted traffic zone. The neural network may output a grid of points representing locations in the environment and information indicating drivability of each respective point. The neural network may output one or more polylines dividing the environment into regions and information indicating whether the AV can drive to or in each respective region.

Abstract: Patterns in lane data and traffic signal data around intersections are used to detect potential errors in map databases and automatically generate data for map databases. Groups of lanes inbound into intersections and/or outbound from intersections can be categorized, and these categories may be used to compare lanes around an intersection, or at multiple intersections along a roadway or in a region. Deviations from expected patterns of these intersection categories may be used to identify errors. Patterns of internal lanes based on external lanes and traffic signal data may be used to automatically generate internal lane data for intersections.

Abstract: Systems and methods for an autonomous vehicle interactive experience platform. An interactive experience platform is provided that utilizes vehicle sensors, as well as in-cabin hardware and software, and fleet management information, to provide a unique and personalized autonomous vehicle interactive experience. The interactive experience platform can use external factors to influence in-experience elements. In some examples, the interactive experience platform can use in-experience events to influence autonomous vehicle behavior.

Abstract: A method is provided for offloading autonomous vehicle (AV) data to a download pole. The AV may include a transceiver. The download pole may include a transceiver. The method may include identifying, by the AV, the download pole in a parking spot that is close to the AV. The method may also include establishing short range wireless link between the AV and the download pole. The method may also include positioning the AV so that the transceiver of the AV is aligned with the transceiver of the download pole based on instructions received over the Bluetooth connection from the download pole. The method may further include establishing a first link between the transceiver of the AV and the transceiver of the download pole.

Abstract: Systems and techniques are provided for validating a simulation framework.

Abstract: Machine learning model optimization systems and methods are disclosed. A system receives sensor data captured by one or more sensors of a vehicle during a first time period. The vehicle uses a first trained machine learning (ML) model for one or more decisions of a first decision type during the first time period. The system generates a second trained ML model at least in part by using the sensor data to train the second trained ML model. The system identifies an optimal trained ML model from a plurality of trained ML models. The plurality of trained ML models includes the first trained ML model and the second trained ML model. The system causes the vehicle to use the optimal trained ML model for one or more further decisions of the first decision type during a second time period after the first time period.

Abstract: The disclosed technology provides solutions for measuring seal characteristics and in particular, for precisely measuring frictional characteristics of a rotary seal. In some aspects, a seal testing apparatus of the disclosed technology can include a motor coupled to a proximal end of a shaft, a rotator coupled to a distal end of the shaft, and a housing disposed around an exterior surface of the rotator, wherein the housing is configured to removably receive at least one seal. In some aspects the seal-testing apparatus can further include a torque transducer coupled to the shaft, wherein the torque transducer is configured to measure a torque output of the motor.