Abstract: Techniques are disclosed for performing hybrid simulation operations with an autonomous vehicle. A method of testing autonomous vehicle operations includes receiving, by a computer, a pre-configured scenario that includes one or more simulation parameters and one or more initial condition parameters, sending, to the autonomous vehicle and based on the one or more initial condition parameters, control signals that instruct the autonomous vehicle to operate at an operative condition, and in response to determining that the autonomous vehicle is operating at the operative condition, performing a simulation with the one or more simulated objects and the autonomous vehicle to test a response of the autonomous vehicle.

Abstract: A system and method for online real-time multi-object tracking is disclosed. A particular embodiment can be configured to: receive image frame data from at least one camera associated with an autonomous vehicle; generate similarity data corresponding to a similarity between object data in a previous image frame compared with object detection results from a current image frame; use the similarity data to generate data association results corresponding to a best matching between the object data in the previous image frame and the object detection results from the current image frame; cause state transitions in finite state machines for each object according to the data association results; and provide as an output object tracking output data corresponding to the states of the finite state machines for each object.

Abstract: Devices, systems and methods for fusing scenes from real-time image feeds from on-vehicle cameras in autonomous vehicles to reduce redundancy of the information processed to enable real-time autonomous operation are described. One example of a method for improving perception in an autonomous vehicle includes receiving a plurality of cropped images, wherein each of the plurality of cropped images comprises one or more bounding boxes that correspond to one or more objects in a corresponding cropped image; identifying, based on the metadata in the plurality of cropped images, a first bounding box in a first cropped image and a second bounding box in a second cropped image, wherein the first and second bounding boxes correspond to a common object; and fusing the metadata corresponding to the common object from the first cropped image and the second cropped image to generate an output result for the common object.

Abstract: Techniques for analysis of autonomous vehicle operations are described. As an example, a method of autonomous vehicle operation includes storing sensor data from one or more sensors located on the autonomous vehicle into a storage medium, performing, based on at least some of the sensor data, a simulated execution of one or more programs associated with the operations of the autonomous vehicle, generating, based on the simulated execution of the one or more programs and as part of a simulation, one or more control signal values that control a simulated driving behavior of a simulated vehicle, and providing a visual feedback of the simulated driving behavior of the simulated vehicle on a simulated road.

Abstract: Techniques are described for estimating pose of a camera located on a vehicle. An exemplary method of estimating camera pose includes obtaining, from a camera located on a vehicle, an image including a lane marker on a road on which the vehicle is driven, and estimating a pose of the camera such that the pose of the camera provides a best match according to a criterion between a first position of the lane marker determined from the image and a second position of the lane marker determined from a stored map of the road.

Abstract: Systems and methods for dynamic predictive control of autonomous vehicles are disclosed. In one aspect, an in-vehicle control system for a semi-truck includes one or more control mechanisms configured to control movement of the semi-truck and a processor. The system further includes computer-readable memory in communication with the processor and having stored thereon computer-executable instructions to cause the processor to receive a desired trajectory and a vehicle status of the semi-truck, determine a dynamic model of the semi-truck based on the desired trajectory and the vehicle status, determine at least one quadratic program (QP) problem based on the dynamic model, generate at least one control command for controlling the semi-truck by solving the at least one QP problem, and provide the at least one control command to the one or more control mechanisms.

Abstract: Techniques are described to determine parameters and/or values for a control model that can be used to operate an autonomous vehicle, such as an autonomous semi-trailer truck. For example, a method of obtaining a data-driven model for autonomous driving may include obtaining data associated with a first set of variables that characterize movements of an autonomous vehicle over time and commands provided to the autonomous vehicle over time, determining, using at least the first set of data, non-zero values and an associated second set of variables that describe a control model used to perform an autonomous driving operation of the autonomous vehicle, and calculating values for a feedback controller that describes a transfer function used to perform the autonomous driving operation of the autonomous vehicle driven on a road.

Abstract: An autonomous vehicle simulation system for analyzing motion planners is disclosed. A particular embodiment includes: receiving map data corresponding to a real world driving environment; obtaining perception data and configuration data including pre-defined parameters and executables defining a specific driving behavior for each of a plurality of simulated dynamic vehicles; generating simulated perception data for each of the plurality of simulated dynamic vehicles based on the map data, the perception data, and the configuration data; receiving vehicle control messages from an autonomous vehicle control system; and simulating the operation and behavior of areal world autonomous vehicle based on the vehicle control messages received from the autonomous vehicle control system.

Abstract: Systems and techniques disclosed in this patent document provide novel and efficient ground transportation of goods by self-driving trucks based on a system of infrastructures with shipping hub facilities serving as warehouses and services centers in accordance with unique characteristics of operating autonomous self-driving trucks without entirely relying on human drivers, including features to cover technological issues, fleet management issues, and customer service issues. In addition, the disclosed systems and techniques may be implemented to integrate other fleet services and modes of transportation of goods with transportation services by self-driving trucks.

Abstract: A system comprises a lead autonomous vehicle (AV), a control device associated with the lead AV, and a following AV. The control device receives a command to navigate the lead AV to avoid an unexpected road condition. The control device receives sensor data from a sensor of the lead AV, comprising location coordinates of objects ahead of the lead AV. The control device accesses environmental data associated with a portion of a road between the lead AV and following AV. The environmental data comprises location coordinates of objects between the lead AV and following AV. The control device determines whether an object in the sensor data or environmental data impedes performing the command by the following AV. The control device updates the command, if the control device determines that an object impedes performing the command by the following AV, and communicates the updated command to the following AV.

Abstract: A system comprises an autonomous vehicle (AV) and an operation server operably coupled with the AV. The operation server accesses environmental data associated with a road traveled by the AV. The environmental data is associated with a time window during which the AV is traveling along the road. The operation server compares the environmental data with map data that comprises expected road conditions ahead of the AV. The operation server determines whether the environmental data comprises an unexpected road condition that is not included in the map data. In response to determining that the environmental data comprises the unexpected road condition that is not included in the map data, the operation server determines a location coordinate of the unexpected road condition, and communicates a command to the AV to maneuver to avoid the unexpected road condition.

Abstract: A system and method for adaptive cruise control with proximate vehicle detection are disclosed. The example embodiment can be configured for: receiving input object data from a subsystem of a host vehicle, the input object data including distance data and velocity data relative to detected target vehicles; detecting the presence of any target vehicles within a sensitive zone in front of the host vehicle, to the left of the host vehicle, and to the right of the host vehicle; determining a relative speed and a separation distance between each of the detected target vehicles relative to the host vehicle; and generating a velocity command to adjust a speed of the host vehicle based on the relative speeds and separation distances between the host vehicle and the detected target vehicles to maintain a safe separation between the host vehicle and the target vehicles.

Abstract: A system and method for vehicle occlusion detection is disclosed.

Abstract: Technique for performing camera calibration on a vehicle is disclosed. A method of performing camera calibration includes emitting, by a laser emitter located on a vehicle and pointed towards a road, a first laser pulse group towards a first location on a road and a second laser pulse group towards a second location on the road, where each laser pulse group includes one or more laser spots. For each laser pulse group: a first set of distances are calculated from a location of a laser receiver to the one or more laser spots, and a second set of distances are determined from an image obtained from a camera, where the second set of distances are from a location of the camera to the one or more laser spots. The method also includes determining two camera calibration parameters of the camera by solving two equations.

Abstract: Disclosed are devices, systems and methods for the operational testing on autonomous vehicles. One exemplary method includes configuring a primary vehicular model with an algorithm, calculating one or more trajectories for each of one or more secondary vehicular models that exclude the algorithm, configuring the one or more secondary vehicular models with a corresponding trajectory of the one or more trajectories, generating an updated algorithm based on running a simulation of the primary vehicular model interacting with the one or more secondary vehicular models that conform to the corresponding trajectory in the simulation, and integrating the updated algorithm into an algorithmic unit of the autonomous vehicle.

Abstract: Various embodiments provide a system and method for iterative lane marking localization that may be utilized by autonomous or semi-autonomous vehicles traveling within the lane. In an embodiment, the system comprises a locating device adapted to determine the vehicle's geographic location; a database; a region map; a response map; a plurality of cameras; and a computer connected to the locating device, database, and cameras, wherein the computer is adapted to receive the region map, wherein the region map corresponds to a specified geographic location; generate the response map by receiving information from the camera, the information relating to the environment in which the vehicle is located; identifying lane markers observed by the camera; and plotting identified lane markers on the response map; compare the response map to the region map; and iteratively generate a predicted vehicle location based on the comparison of the response map and the region map.

Abstract: A system comprises an autonomous vehicle (AV) and a control device operably coupled with the AV. The control device detects a series of events within a threshold period of time, where a number of series of events in the series of events is above a threshold number. The series of events taken in the aggregate within the threshold period of time deviates from a normalcy mode. The normalcy mode comprises events that are expected to the encountered by the AV. The control device determines whether the series of events corresponds to a malicious event, where the malicious event indicates tampering with the AV. In response to determining that the series of events corresponds to the malicious event, the series of events are escalated to be addressed.

Abstract: A system and method for taillight signal recognition using a convolutional neural network is disclosed. An example embodiment includes: receiving a plurality of image frames from one or more image-generating devices of an autonomous vehicle; using a single-frame taillight illumination status annotation dataset and a single-frame taillight mask dataset to recognize a taillight illumination status of a proximate vehicle identified in an image frame of the plurality of image frames, the single-frame taillight illumination status annotation dataset including one or more taillight illumination status conditions of a right or left vehicle taillight signal, the single-frame taillight mask dataset including annotations to isolate a taillight region of a vehicle; and using a multi-frame taillight illumination status dataset to recognize a taillight illumination status of the proximate vehicle in multiple image frames of the plurality of image frames, the multiple image frames being in temporal succession.

Abstract: Disclosed are devices and systems for an optimized sensor layout for an autonomous or semi-autonomous vehicle. In one aspect, the system includes a vehicle capable of semi-autonomous or autonomous operation. A plurality of forward-facing cameras is coupled to the vehicle and configured to have a field of view in front of the vehicle. At least three forward-facing cameras of the plurality of forward-facing cameras have different focal lengths. A right-side camera is coupled to the right side of the vehicle, the right-side camera configured to have a field of view to the right of the vehicle. A left-side camera is coupled to the left side of the vehicle, the left-side camera is configured to have a field of view to the left of the vehicle.

Abstract: A system and method for three-dimensional (3D) object detection is disclosed. A particular embodiment can be configured to: receive image data from a camera associated with a vehicle, the image data representing an image frame; use a machine learning module to determine at least one pixel coordinate of a two-dimensional (2D) bounding box around an object in the image frame; use the machine learning module to determine at least one vertex of a three-dimensional (3D) bounding box around the object; obtain camera calibration information associated with the camera; and determine 3D attributes of the object using the 3D bounding box and the camera calibration information.