Abstract: The present disclosure is directed generating candidate trajectories and selecting one of them to be implemented. In particular, a computing system can obtain an initial travel path for an autonomous vehicle. The computing system can obtain sensor data describing objects within an environment of the autonomous vehicle. The computing system can generate a plurality of trajectories for the autonomous vehicle based on the sensor data and the initial travel path. The computing system can determine whether the initial travel path, an offset profile, and a velocity profile meet flatness criteria. In response to determining that the initial travel path, the offset profile, and the velocity profile meet the flatness criteria, the computing system can combine the initial travel path, the offset profile, and the velocity profile into the respective trajectory. The computing system can determine a trajectory from the plurality of trajectories.

Abstract: A control system of a self-driving tractor can access sensor data to determine a set of trailer configuration parameters of a cargo trailer coupled to the self-driving tractor. Based on the set of trailer configuration parameters, the control system can configure a motion planning model for autonomously controlling the acceleration, braking, and steering systems of the tractor.

Abstract: Systems, methods, tangible non-transitory computer-readable media, and devices associated with vehicle control based on risk-based interactions are provided. For example, vehicle data and perception data can be accessed. The vehicle data can include the speed of an autonomous vehicle in an environment. The perception data can include location information and classification information associated with an object in the environment. A scenario exposure can be determined based on the vehicle data and perception data. Prediction data including predicted trajectories of the object can be accessed. Expected speed data can be determined based on hypothetical speeds and hypothetical distances between the vehicle and the object. A speed profile that satisfies a threshold criteria can be determining based on the scenario exposure, the prediction data, and the expected speed data, over a distance. A motion plan to control the autonomous vehicle can be generated based on the speed profile.

Abstract: Systems and methods for controlling autonomous vehicles are provided. In one example embodiment, a computing system can obtain data indicative of a service assignment associated with an autonomous vehicle. The service assignment is indicative of a destination location to which to transport a user. The computing system can obtain data indicative of an item associated with the user that is to be transported to the destination location. The computing system can determine a first drop-off location for the user and a second drop-off location for the item based at least in part on the destination location. The first drop-off location for the user is different than the second drop-off location for the item. The computing system can cause the autonomous vehicle to initiate motion control to travel to the first drop-off location for the user and to travel to the second drop-off location for the item.

Abstract: Systems and methods are directed to detecting tampering with a secured area of an autonomous vehicle. In one example, a computing system performs operations including generating a first key pair comprising a first private key and a first public key as part of a vehicle registration process; providing the first public key to one or more remote computing systems; determining whether a reset event occurred at a vehicle; in response to determining that the reset event occurred, discarding the first key pair and generating a second key pair comprising a second private key and a second public key; receiving a status query from one of the one or more remote computing systems; generating a response to the status query, the response being signed using either the first private key or the second private key; and providing the response to the one of the one or more remote computing systems.

Abstract: Systems and methods for controlling autonomous vehicles are provided. In one example embodiment, a computer-implemented method includes obtaining, by a computing system that includes one or more computing devices onboard an autonomous vehicle, data indicative of an identifier associated with a user device of a user. The identifier can be assigned to the user device by a cellular network. The method includes determining, by the computing system, location data associated with the user device based at least in part on the identifier associated with the user device. The method includes determining, by the computing system, a vehicle route based at least in part on the location data associated with the user device. The method includes causing, by the computing system, the autonomous vehicle to initiate travel in accordance with the vehicle route.

Abstract: Systems, methods, tangible non-transitory computer-readable media, and devices associated with the operation of a vehicle are provided. For example a vehicle computing system can receive occupancy data that includes information associated with occupancy of a vehicle that includes seats. One or more states of the vehicle can be determined. The states of the vehicle can include a disposition of any object that is within the vehicle. Further, a configuration of the seats in the vehicle can be determined based on the occupancy data and the states of the vehicle. The configuration can include a disposition of the seats inside the vehicle. Furthermore, at least one of the seats can be adjusted based on the configuration that was determined.

Abstract: Generally, the disclosed systems and methods utilize multi-task machine-learned models for object intention determination in autonomous driving applications. For example, a computing system can receive sensor data obtained relative to an autonomous vehicle and map data associated with a surrounding geographic environment of the autonomous vehicle. The sensor data and map data can be provided as input to a machine-learned intent model. The computing system can receive a jointly determined prediction from the machine-learned intent model for multiple outputs including at least one detection output indicative of one or more objects detected within the surrounding environment of the autonomous vehicle, a first corresponding forecasting output descriptive of a trajectory indicative of an expected path of the one or more objects towards a goal location, and/or a second corresponding forecasting output descriptive of a discrete behavior intention determined from a predefined group of possible behavior intentions.

Abstract: Systems and method for video compression using conditional entropy coding. An ordered sequence of image frames can be transformed to produce an entropy coding for each image frame. Each of the entropy codings provide a compressed form of image information based on a prior image frame and a current image frame (the current image frame occurring after the prior image frame). In this manner, the compression model can capture temporal relationships between image frames or encoded representations of the image frames using a conditional entropy encoder trained to approximate the joint entropy between frames in the image frame sequence.

Abstract: Systems and methods are directed to deploying warning devices by an autonomous vehicle. In one example, a system includes one or more processors and memory including instructions that, when executed by the one or more processors, cause the one or more processors to perform operations. The operations include obtaining data indicating that a vehicle stop maneuver is to be implemented for an autonomous vehicle. The operations further include determining whether one or more warning devices should be dispensed from the autonomous vehicle during the vehicle stop maneuver based in part on the obtained data. The operations further include, in response to determining one or more warning devices should be dispensed from the autonomous vehicle, determining a dispensing maneuver for the one or more warning devices, and providing one or more command signals to one or more vehicle systems to perform the dispensing maneuver for the one or more warning devices.

Abstract: Systems, methods, tangible non-transitory computer-readable media, and devices associated with the operation of a vehicle are provided. For example, a wheel unit system can receive vehicle data including information associated with states of a vehicle. Further, the vehicle can include wheel wells associated with wheel units that are configured to generate indications associated with the states of the vehicle. The wheel unit system can also determine indications to generate on the wheel units based at least in part on the vehicle data. Furthermore, the wheel unit system can generate the indications on the wheel units. The indications can, for example, include visual indications or auditory indications.

Abstract: Systems and methods for determining a location based on image data are provided. A method can include receiving, by a computing system, a query image depicting a surrounding environment of a vehicle. The query image can be input into a machine-learned image embedding model and a machine-learned feature extraction model to obtain a query embedding and a query feature representation, respectively. The method can include identifying a subset of candidate embeddings that have embeddings similar to the query embedding. The method can include obtaining a respective feature representation for each image associated with the subset of candidate embeddings. The method can include determining a set of relative displacements between each image associated with the subset of candidate embeddings and the query image and determining a localized state of a vehicle based at least in part on the set of relative displacements.

Abstract: Systems, methods, tangible non-transitory computer-readable media, and devices associated with object association and tracking are provided. Input data can be obtained. The input data can be indicative of a detected object within a surrounding environment of an autonomous vehicle and an initial object classification of the detected object at an initial time interval and object tracks at time intervals preceding the initial time interval. Association data can be generated based on the input data and a machine-learned model. The association data can indicate whether the detected object is associated with at least one of the object tracks. An object classification probability distribution can be determined based on the association data. The object classification probability distribution can indicate a probability that the detected object is associated with each respective object classification. The association data and the object classification probability distribution for the detected object can be outputted.

Abstract: Various examples are directed to a rotary coupler and methods of use thereof. The rotary data coupler may comprise a transmitter and receiver. The transmitter may comprise a first band and a second transmitter band. The receiver may comprise a receiver housing positioned to rotate relative to the first transmitter band and the second transmitter band. A first receiver band may be positioned opposite the first transmitter band to form a first capacitor and a second receiver band may be positioned opposite the second transmitter band to form a second capacitor. The receiver may also comprise a resistance electrically coupled between the first receiver band and the second receiver band and a differential amplifier. The differential amplifier may comprise an inverting input and a non-inverting input, with the non-inverting input electrically coupled to the first receiver band and the inverting input electrically coupled to the second receiver band.

Abstract: A method includes obtaining map data associated with a map of a geographic location including one or more roadways, the map including a first submap represented by a first local Euclidean space and a second submap represented by a second local Euclidean space. A route that includes a first roadway in the first submap and a second roadway in the second submap is determined using a first projection between a global coordinate system and the first local Euclidean space and a second projection between the global coordinate system and the second local Euclidean space. The route is provided to an autonomous vehicle (AV) for driving on the first roadway and the second roadway.

Abstract: In one example embodiment, a computer-implemented method for autonomous vehicle control includes determining whether a cargo container is attached to the vehicle base. The method includes controlling a front shield associated with an autonomous vehicle to move from a closed position to an opened position when a cargo container is determined to be attached to a vehicle base associated with the autonomous vehicle. The method includes controlling the front shield to move from the opened position to the closed position when the cargo container is not attached to the vehicle base.

Abstract: Systems and methods for determining object intentions through visual attributes are provided. A method can include determining, by a computing system, one or more regions of interest. The regions of interest can be associated with surrounding environment of a first vehicle. The method can include determining, by a computing system, spatial features and temporal features associated with the regions of interest. The spatial features can be indicative of a vehicle orientation associated with a vehicle of interest. The temporal features can be indicative of a semantic state associated with signal lights of the vehicle of interest. The method can include determining, by the computing system, a vehicle intention. The vehicle intention can be based on the spatial and temporal features. The method can include initiating, by the computing system, an action. The action can be based on the vehicle intention.

Abstract: An autonomous robot is provided. In one example embodiment, an autonomous robot can include a main body including one or more compartments. The one or more compartments can be configured to provide support for transporting an item. The autonomous robot can include a mobility assembly affixed to the main body and a sensor configured to obtain sensor data associated with a surrounding environment of the autonomous robot. The autonomous robot can include a computing system configured to plan a motion of the autonomous robot based at least in part on the sensor data. The computing system can be operably connected to the mobility assembly for controlling a motion of the autonomous robot. The autonomous robot can include a coupling assembly configured to temporarily secure the autonomous robot to an autonomous vehicle. The autonomous robot can include a power system and a ventilation system that can interface with the autonomous vehicle.

Abstract: A georeferenced trajectory estimation system receives data generated by vehicle sensors from a fleet of vehicles and identifies sets of sensor measurements for a geographic region taken at different times. The sets of sensor measurements include environment geometry data and unaligned trajectory data in local coordinate frames. The georeferenced trajectory estimation system aligns the environment geometry data between sets of sensor measures, and, based on the alignment of the environment geometry data, transforms the identified corresponding trajectory data into a common coordinate frame.

Abstract: Aspects of the present disclosure involve systems, methods, and devices for determining specular reflectivity characteristics of objects using a Lidar system of an autonomous vehicle (AV) system. A method includes transmitting at least two light signals directed at a target object utilizing the Lidar system of the AV system. The method further includes determining at least two reflectivity values for the target object based on return signals corresponding to the at least two light signals. The method further includes classifying specular reflectivity characteristics of the target object based on a comparison of the first and second reflectivity value. The method further includes updating a motion plan for the AV system based on the specular reflectivity characteristics of the target object.