Abstract: Methods and systems for controlling navigation of a vehicle are disclosed. The system will first detect a URU within a threshold distance of a drivable area that a vehicle is traversing or will traverse. The system will then receive perception information relating to the URU, and use a plurality of features associated with each of a plurality of entry points on a drivable area boundary that the URU can use to enter the drivable area to determine a likelihood that the URU will enter the drivable area from that entry point. The system will then generate a trajectory of the URU using the plurality of entry points and the corresponding likelihoods, and control navigation of the vehicle while traversing the drivable area to avoid collision with the URU.

Abstract: This document discloses system, method, and computer program product embodiments for operating an autonomous vehicle (AV). For example, the method includes performing the following operations by a muxing tool when AV is deployed within a particular geographic area in a real-world environment: receiving perception data that is representative of at least one actual object which is perceived while AV is deployed within the particular geographic area in a real-world environment; receiving simulation data that represents a simulated object that could be perceived by AV in the real-world environment and that was generated using a simulation scenario which is selected from a plurality of simulation scenarios based on at least one of the particular geographic area in which AV is currently located and a current operational state of AV; and generating augmented perception data by combining the simulation data with the perception data.

Abstract: Systems/methods for operating an autonomous vehicle. The methods comprise, by a computing device: using sensor data to track an object that was detected in proximity to the autonomous vehicle; classifying the object into at least one dynamic state class of a plurality of dynamic state classes; transforming the at least one dynamic state class into at least one goal class of a plurality of goal classes; transforming the at least one goal class into at least one proposed future intention class of a plurality of proposed future intention classes; determining at least one predicted future intention of the object based on the proposed future intention class; and/or causing the autonomous vehicle to perform at least one autonomous driving operation based on the at least one predicted future intention determined for the object.

Abstract: Systems, methods, and autonomous vehicles may obtain one or more images associated with an environment surrounding an autonomous vehicle; determine, based on the one or more images, an orientation of a head worn item of protective equipment of an operator of a vehicle; determine, based on the orientation of the head worn item of protective equipment, a direction of a gaze of the operator and a time period associated with the direction of the gaze of the operator; determine, based on the direction of the gaze of the operator and the time period associated with the direction of the gaze of the operator, a predicted motion path of the vehicle; and control, based on the predicted motion path of the vehicle, at least one autonomous driving operation of the autonomous vehicle.

Abstract: A relative position of an object in the surroundings of a vehicle is estimated based on a two-dimensional camera image. A control unit determines an object contour of the object from the camera image and determines at least one digital object template that represents the object based on the object contour. The control unit forward projects the at least one object template from respective different positions onto an image plane of the camera image. Each forward-projected object template yields a respective two-dimensional contour proposal, and the control unit compares the contour proposals with the object contour of the object.

Abstract: Disclosed herein are system, method, and computer program product embodiments for switching between local and remote guidance instructions for autonomous vehicles. For example, the method includes, in response to monitoring one or more actions of objects detected in a scene in which the autonomous robotic system is moving, causing the autonomous robotic system to slow or cease movement in the scene. The method includes detecting a trigger condition based on movement of the autonomous robotic system in the scene. In response to the one or more monitored actions and detecting the trigger condition, the method includes transmitting a remote guidance request to a remote server. After transmitting the remote guidance request, the method includes receiving remote guidance instructions from the remote server and causing the autonomous robotic system to begin operating according to the remote guidance instructions.

Abstract: Methods and systems for controlling navigation of an autonomous vehicle for traversing a drivable area are disclosed. The methods include receiving information relating to a drivable area that includes a plurality of polygons, identifying a plurality of logical edges that form a boundary of the drivable area, sequentially and repeatedly analyzing concavities of each the plurality of logical edges until identification of a first logical edge that has a concavity greater than a threshold, creating a first logical segment of the boundary of the drivable area. This segmentation may be repeated until each of the plurality of logical edges has been classified. The method may include creating and adding (to a map) a data representation of the drivable area that comprises an indication of the plurality of logical segments, and adding the data representation to a road network map comprising the drivable area.

Abstract: Disclosed herein are system, method, and computer program product embodiments for automated autonomous vehicle pose validation. An embodiment operates by generating a range image from a point cloud solution comprising a pose estimate for an autonomous vehicle. The embodiment queries the range image for predicted ranges and predicted class labels corresponding to lidar beams projected into the range image. The embodiment generates a vector of features from the range image. The embodiment compares a plurality of values to the vector of features using a binary classifier. The embodiment validates the autonomous vehicle pose based on the comparison of the plurality of values to the vector of features using the binary classifier.

Abstract: A time master and sensor data collection module for a robotic system such as an autonomous vehicle is disclosed. The module includes a processing device, one or more sensors, and programming instructions that are configured to cause the processing device to operate as a timer that generates a vehicle time, receive data from the one or more sensors contained within the housing, and synchronize the data from the one or more sensors contained within the housing with the vehicle time. The integrated sensors may include sensors such as a global positioning system (GPS) unit and/or an inertial measurement unit (IMU). The module may interface with external sensors such as a LiDAR system and/or cameras.

Abstract: Disclosed herein are system, method, and computer program product embodiments for clustering lane segments of a roadway in order to improve and simplify autonomous vehicle behavior testing. The approaches disclosed herein provide a hybrid methodology of dividing lane segments into hard features and soft features, and using a metric learning model trained in a supervised process on the entirety of lane segment features to cluster the lane segments based on the soft features. These clustered lane segments can then be assigned to what is termed as protolanes, where a single set of tests applied to a given protolane is considered valid across all of the lane segments assigned to the protolane.

Abstract: Disclosed herein are system, method, and computer program product aspects for testing autonomous vehicle intent using protolanes. For example, the method includes selecting a first lane segment of a roadway within a geonet, wherein the geonet comprises a plurality of lane segments. It is determined that the first lane segment corresponds to a protolane of a set of protolanes, wherein the protolane is associated with one or more additional lane segments of the plurality of lane segments. The protolane is associated with the first lane segment based on the determination that the first lane segment corresponds to the protolane.

Abstract: A system receives a road network map that corresponds to a road network that is in an environment of an autonomous vehicle. For each of the one or more lane segments, the system identifies one or more conflicting lane segments from the plurality of lane segments, each of which conflicts with the lane segment, and adds conflict data pertaining to a conflict between the lane segment and the one or more conflicting lane segments to a set of conflict data. The system analyzes the conflict data to identify a conflict cluster that is representative of an intersection. The system groups predecessor lane segments and the successor lane segments as inlets or outlets of the intersection, generates an outer geometric boundary of the intersection, generates an inner geometric boundary of the intersection, creates a data representation of the intersection and adds the data representation to the road network map.

Abstract: Systems and methods for managing data. The methods comprise: receiving, by a load driver node of a network, a request for upload of source data associated with a resource name to a data warehouse; generating a publication identifier and a version value for the source data by the load driver node in response to the request; causing, by the load driver node, operations to be performed by load worker nodes to facilitate population of at least one fact table of the data warehouse with the publication identifier and the source data; and causing, by the load driver node, a publication table of the data warehouse to be updated to include the publication identifier and the version value so as to be associated with the resource name.

Abstract: Systems, methods, and autonomous vehicles for detecting road marking points from LiDAR data may obtain a LiDAR dataset generated by a LiDAR system; process, for each laser emitter of the LiDAR system, a point cloud associated with that laser emitter by: determining, for each point in the point cloud, based on a gradient of intensity at that point, whether that point corresponds to a gradient edge point; and determining, based on intensities of points in an interior segment of points between at least one candidate pair of gradient edge points and intensities of points in exterior segments of points outside the at least one candidate pair of gradient edge points, whether that candidate pair of gradient edge points corresponds to a road marking edge; and aggregate, from the processing of the plurality of point clouds, candidate pairs of gradient edge points determined to correspond to road marking edges.

Abstract: Systems, methods, and computer-readable media are disclosed for obtaining data from multiple internal vehicle networks. An example method may include receiving, using a first internal communication network of a vehicle, a first query for data from one or more devices of the vehicle, wherein the first query is formatted for communication over the first internal communication network. The example method may also include generating, based on the first query, a second query associated with a supplemental communication network, wherein the supplemental communication network is also internal to the vehicle. The example method may also include sending, using the supplemental communication network, the second query to a first device, the first device being located on the supplemental communication network. The example method may also include receiving, from the first device, data relating to the second query.

Abstract: A mounting device for selectively positioning an optical element within a field-of-view of an optical sensor of a vehicle includes: a housing defining an opening sized to fit over an aperture of the optical sensor; a holder for the optical element connected to the housing and positioned such that, when the holder is in a first position, the optical element is at least partially within the field-of-view of the optical sensor; and a motorized actuator. The motorized actuator can be configured to move the holder to adjust the position of the optical element relative to the field-of-view of the optical sensor.

Abstract: Devices, systems, and methods are provided for communications between autonomous and emergency vehicles. A method may include identifying, by an autonomous vehicle (AV), a first message received from a first vehicle, and identifying, by the AV, in the first message, information associated with identifying the AV, a security key associated with identifying the first vehicle, and an instruction associated with causing the AV to perform an action. The method may include authenticating, by the AV, based on the security key, the first vehicle, and controlling operation, based on the instruction and the information associated with identifying the AV, of the AV to perform the action.

Abstract: Systems and methods for selecting data for training a machine learning model using active learning are disclosed. The methods include receiving a plurality of unlabeled sensor data logs corresponding to surroundings of an autonomous vehicle, identifying one or more trends associated with a training dataset comprising a plurality of labeled data logs, determining a function for assigning an importance score to each of the plurality of unlabeled sensor data logs, using the one or more trends, using the function for assigning the importance score to each of the plurality of unlabeled sensor data logs, selecting a subset of the plurality of sensor data logs that have an importance score greater than a threshold, and using the subset of the plurality of sensor data logs for further training the machine learning model trained using the training dataset to generate an updated model.

Abstract: Systems, methods, and computer-readable media are disclosed for a systems and methods for improved LIDAR return light capture efficiency. One example method may include comparing, by a controller including a processor and at a first time, a first temperature of a first computing element to a first threshold temperature and a second temperature of a second computing element to a second threshold temperature. The example method may also include sending, based on a determination that the first temperature is below the first threshold temperature and the second temperature is above the second threshold temperature, a first signal to a switch to activate a data output corresponding to the second computing element. The example method may also include sending, to the second computing element, a second signal to cause a third computing element to increase heat dissipation from the third computing element to the first computing element.

Abstract: Methods of forecasting intentions of actors that an autonomous vehicle (AV) encounters in are disclosed. The AV uses the intentions to improve its ability to predict trajectories for the actors, and accordingly making decisions about its own trajectories to avoid conflict with the actors. To do this, for any given actor the AV determines a class of the actor and detects an action that the actor is taking. The system uses the class and action to identify candidate intentions of the actor and evaluating a likelihood of each candidate intention. The system repeats this process over multiple cycles to determine overall probabilities for each of the candidate intentions. The AV's motion planning system can use the probabilities to determine likely trajectories of the actor, and accordingly influence the trajectory that the AV will itself follow in the environment.