Abstract: Categorizing driving behaviors of other road users includes maintaining a first history of first lateral-offset values of a road user with respect to a center line of a lane of a road; determining a first pattern based on the first history of the first lateral-offset values; determining a driving behavior of the road user based on the first pattern; and autonomously performing, by a host vehicle, a driving maneuver based on the driving behavior of the road user. The first history can be maintained for a predetermined period of time. An apparatus includes a processor that is configured to track a trajectory history of a road user; determine, based on the trajectory history, a driving behavior of the road user; and transmit a notification of the driving behavior.

Abstract: A visibility grid is determined for at least one external object within an operational environment of an autonomous vehicle. The visibility grid represents areas visible and areas obstructed for the external object. When the autonomous vehicle is positioned within an occluded region relative to the visibility grid of the external object is identified. Driving parameters of the autonomous vehicle are altered in response to being in the occluded region to avoid collision with the external object.

Abstract: At least one virtual road user is generated, wherein a position of a respective virtual road user of the at least one virtual road user corresponds to a border of a range of a sensor of a host vehicle approaching an intersection of a vehicle transportation network. A most relevant virtual road user of the at least one virtual road user is determined, the most relevant virtual road user being associated with an earliest crossing lane of the intersection from a perspective of the host vehicle. A time to contact for the most relevant virtual road user is determined, wherein the time to contact is based on an acceleration of the host vehicle, a predicted trajectory of the most relevant virtual road user, and a relative distance between the host vehicle and the most relevant virtual road user. A target speed for the host vehicle is determined based on the time to contact and the relative distance. The host vehicle is operated using the target speed as input to a control system of the host vehicle.

Abstract: Observed driveline mean and variance data are used for determining the variance of a trajectory of tracked objects for use by a host vehicle. A map of a portion of a vehicle transportation network are determined, wherein the map is comprised of observed driveline mean and variance data for one or more map points. At least one trajectory of a tracked object is predicted, wherein a trajectory includes a series of location each corresponding to a respective predicted position of the tracked object at a future time. A map-based variance is generated for the location of the trajectory using a smoothed curvature of the trajectory within the map. A control system of the vehicle operates the vehicle using the map-based variance as input.

Abstract: Determining a speed plan for an autonomous vehicle (AV) is disclosed. Planned locations of the AV for future time steps are placed in an occupancy grid. The planned locations are based on a strategic speed plan that is determined without taking world objects into account. Predicted locations of the world objects for at least some of the future time steps are placed in the occupancy grid. Respective buffer distances corresponding to the predicted locations are added in the occupancy grid. An estimated speed plan is identified for the AV based on the occupancy grid. The speed plan is obtained from the estimated speed plan. The AV is then controlled according to the speed plan.

Abstract: A system may receive a video stream from a camera of a vehicle in a transportation network. The system may also receive an input indicating an acceleration and a steering angle of a simulated lead vehicle. The system may determine a pose of the simulated vehicle relative to a home pose based on the acceleration input and the steering input, and display an overlay of a representation of the simulated vehicle in the video stream at pixel coordinates based on the pose. The system may determine, from the pixel coordinates, spatial coordinates of the simulated vehicle in the transportation network, wherein the spatial coordinates are relative to at least one of the transportation network or the camera. The system may transmit the spatial coordinates to the vehicle to cause the vehicle to follow a path based on the spatial coordinates.

Abstract: Map and kinematic based predictions are used for determining the trajectory of road users for use by a host vehicle. A kinematic trajectory of a road user is determined. The road user is associated with mapped lanes. At least one path of the road user is predicted using the mapped lanes. For a path of the at least one path, a probability is generated using the kinematic trajectory of the road user, wherein the probability represents how likely the road user will continue following the path. A kinematic prediction of the road user is generated using the probability corresponding to the path, wherein the kinematic prediction represents how likely the road user will follow an alternative path to the at least one path. Using at least one control system of a vehicle, a control action for the vehicle is determined using the at least one path and the kinematic prediction.

Abstract: Point cloud data from a sensor of a vehicle at different distances ahead of the vehicle are accumulated over time. Density data of the point cloud data at the different distances ahead of the vehicle are identified. A road irregularity is identified based on the density data. The vehicle is controlled in response to the irregularity.

Abstract: A data-based driveline map is determined using road-level map data and driveline data for road users. The map data includes way data comprising one or more ways, a way includes a series of nodes, and the driveline data includes drivelines, where a driveline is a series of poses representing a road user. A first section of the way data is identified as an intersection, and second sections are matched with the driveline data to generate multiple way bars, where a way bar includes one or more poses and one node. A way bar is categorized as either constant or changing based on lanes counted therewithin. Consecutive way bars are grouped into way bar sections based on the categorization and the lane count, and the map is generated using the way bar sections and the first section. A vehicle is operated using the map as input to a control system.

Abstract: A mean of real-time accelerator pedal output of a vehicle that quantifies an extent to which an accelerator pedal has been pressed by a driver of the vehicle over a defined period of time is determined. Target mean accelerator pedal output for the vehicle is determined. Torque of the vehicle is changed. The torque is reduced when the mean of the real-time accelerator pedal output is lower than the target mean accelerator pedal output, and the torque is increased when the mean of the real-time accelerator pedal output is higher than the target mean accelerator pedal output.

Abstract: An occlusion is identified in a vehicle transportation network. A visibility grid is identified on a second side of the occlusion for a vehicle that is on a first side of the occlusion. The visibility grid is identified with respect to a region of interest that is at least a predefined distance above ground. The visibility grid is used to identify first portions of roads sensed by a sensor positioned on the vehicle and second portions of the roads that are not sensed by the sensor. A driving behavior of the vehicle is altered based on the visibility grid.

Abstract: Proactively mitigating risk to a vehicle traversing a vehicle transportation network includes identifying a location for a virtual vehicle. The virtual vehicle is added to a world object model maintained with respect to the vehicle. A trajectory is predicted for the virtual vehicle. The vehicle is autonomously controlled according to an adjusted trajectory that is based on the trajectory for the virtual vehicle. The adjusted trajectory includes at least one of a lateral constraint or a speed constraint. The location for the virtual vehicle is identified based on a lane in map data, a trajectory of a vehicle, and a perceptible area by sensors of the vehicle. The virtual vehicle is a hypothetical vehicle that is not observed by sensors of the vehicle.

Abstract: A system can determine, from a vehicle traversing in a vehicle transportation network, movement information associated with an object traveling in front of the vehicle and road information associated with the vehicle transportation network. The system can then determine a probability of the object representing a backup or stopping (BoS) hazard to the vehicle. The probability can be based on the movement information and the road information. The system can then assign a BoS classification to the object based on the probability exceeding a threshold. The system can then calculate, based on assigning the BoS classification, a risk zone representing a target minimum separation distance between the vehicle and the object. The system can then control the vehicle to avoid the risk zone by constraining a speed of the vehicle.

Abstract: Detecting prediction errors includes detecting a road user; determining respective predicted data for the road user; storing, in a data structure, the respective predicted data; storing, in the data structure, actual data of the road user; obtaining an average prediction displacement error using at least one of the actual data and at least two corresponding respective predicted data; and determining a prediction accuracy based on the average prediction displacement error. Detecting map errors includes detecting a road user; storing, in a data structure, actual data of the road user; storing, in the data structure, map data corresponding to the actual data; obtaining an average map displacement error based on a comparison of at least some of the actual data and corresponding at least some map data; and determining a map accuracy based on the average map displacement error.

Abstract: Proactively mitigating risk to a vehicle traversing a vehicle transportation network is described. First and second hazard zones for first and second objects ahead of the vehicle are respectively determined. The first hazard zone includes a first target lateral constraint that extends over a left lane boundary, and the second hazard zone includes a second target lateral constraint that extends over a right lane boundary. The lateral constraints separately allow the vehicle to avoid the objects without a speed constraint. Where the first and second hazard zones overlap in the longitudinal direction, a lateral buffer is allocated between the lateral constraints to generate first and second allocated lateral constraints. Longitudinal constraints are respectively determined based on times of arrival at each hazard zone. Using the constraints, a proactive trajectory is determined that includes a lateral contingency, a longitudinal contingency, or both. The vehicle is controlled according to the trajectory.

Abstract: Object avoidance by an autonomous vehicle (AV) is disclosed. A method includes detecting a first object along a coarse driveline of a drivable area of the AV; receiving a predicted path of the first object; determining, based on the predicted path of the first object, an adjusted drivable area; and determining a trajectory of the AV through the adjusted drivable area. A system includes a trajectory planner configured to detect a first object along a coarse driveline of a drivable area of the AV; receive a predicted path of the first object; determine, based on the predicted path of the first object, an adjusted drivable area; and determine a trajectory of the AV through the adjusted drivable area.

Abstract: Trajectory planning include identifying state data based on sensor data from sensors of a vehicle. The state data are input to a machine-learning model to obtain parameters of a trajectory planner. The parameters are input to the trajectory planner to obtain a short term speed plan. The vehicle is autonomously controlled according to at least a portion of the short term speed plan.

Abstract: Engine activation planning in a hybrid electric vehicle (HEV) is disclosed. Historical driving data of the HEV are analyzed to identify recurring driving scenarios and patterns specific to a driver of the HEV. Future driving scenarios are predicted based on the historical driving data. An engine activation policy is generated for the HEV. The engine activation policy optimizes battery charging and usage in response to the future driving scenarios. The engine activation policy is used to control activation of a gasoline engine in the HEV where a control decision is dynamically adjusted based on a comparison of real-time driving conditions with the future driving scenarios.

Abstract: A vehicle includes a vehicle engine, a steering control unit, an on-board sensor network and a navigational constraint control system. The vehicle engine generates a torque output of the vehicle. The steering control unit controls a steering angle of the vehicle. The on-board sensor network is programmed to detect external objects within a detection zone. The navigational constraint control system has a memory for storing a path index for the vehicle's navigation. The processor is programmed to determine a reference trajectory from the path index. The processor is further programmed to calculate navigational constraints for the determined reference trajectory to determine a nominal trajectory based on information detected by the on-board sensor network. The processor is programmed to control at least one of the vehicle engine and the steering control unit in accordance with the nominal trajectory.

Abstract: A vehicle includes a vehicle engine, a steering control unit, an on-board sensor network and a virtual hazard inference system. The vehicle engine generates a torque output of the vehicle. The steering control unit controls a steering angle of the vehicle. The on-board sensor network is programmed to detect external objects within a detection zone. The virtual hazard inference system has a processor programmed to infer a presence of an occluded object that is not detected by the on-board sensor network. The processor controls the vehicle engine and the steering control unit to perform at least one of a torque control operation and a steering control operation.