Abstract: Devices, systems, and methods for remote control of autonomous vehicles are disclosed herein. A method may include receiving, by a device, first data indicative of an autonomous vehicle in a parking area, and determining, based on the first data, a location of the autonomous vehicle. The method may include determining, based on a the location, first image data including a representation of an object. The method may include generating second image data based on the first data and the first image data, and presenting the second image data. The method may include receiving an input associated with controlling operation of the autonomous vehicle, and controlling, based on the input, the operation of the autonomous vehicle.

Abstract: The present disclosure relates generally to identification of drivable surfaces in connection with autonomously performing various tasks at industrial work sites and, more particularly, to techniques for distinguishing drivable surfaces from non-drivable surfaces based on sensor data. A framework for the identification of drivable surfaces is provided for an autonomous machine to facilitate it to autonomously detect the presence of a drivable surface and to estimate, based on sensor data, attributes of the drivable surface such as road condition, road curvature, degree of inclination or declination, and the like. In certain embodiments, at least one camera image is processed to extract a set features from which surfaces and objects in a physical environment are identified, and to generate additional images for further processing. The additional images are combined with a 3D representation, derived from LIDAR or radar data, to generate an output representation indicating a drivable surface.

Abstract: Methods and systems for jointly estimating a pose and a shape of an object perceived by an autonomous vehicle are described. The system includes data and program code collectively defining a neural network which has been trained to jointly estimate a pose and a shape of a plurality of objects from incomplete point cloud data. The neural network includes a trained shared encoder neural network, a trained pose decoder neural network, and a trained shape decoder neural network. The method includes receiving an incomplete point cloud representation of an object, inputting the point cloud data into the trained shared encoder, outputting a code representative of the point cloud data. The method also includes generating an estimated pose and shape of the object based on the code. The pose includes at least a heading or a translation and the shape includes a denser point cloud representation of the object.

Abstract: A method and system allowing fully autonomous automatic take-off using only images captured by cameras on the aircraft and avionics data. The system includes an image capture device on the aircraft to take a stream of images of the runway, image processing modules to estimate, on the basis of the streams of images, a preliminary current position of the aircraft on the runway and to assign a preliminary confidence index to the estimate. A data consolidation module can determine a relevant current position of the aircraft on the runway by consolidating data originating from the image processing modules with inertial data to correct the estimate of the preliminary current position and determine a relevant confidence index using a current speed of the wheels of the aircraft to refine the preliminary confidence index. A flight control computer can control and guide aircraft take-off.

Abstract: A system includes a plurality of sensors along a surface of an airfoil operable to measure a first set of ice thickness values at a first time and a second set of ice thickness values at a second time. The system further includes a processor configured to determine a first plurality of lift calculation variables and a second plurality of lift calculation variables. The processor also generates a threshold angle of attack value and updates the threshold angle of attack value at the second time, based on one or more differences between the first and second sets of ice thickness values and the first and second plurality of lift calculation variables. The processor is further configured to send, to a display, based on the updated threshold angle of attack, one or more changes to flight data to adjust the actual angle of attack of the airfoil.

Abstract: A method for training a machine learning model, e.g., a neural network, using a regularization scheme is disclosed. The method includes generating regularized partial gradients of losses computed using an objective function for training the machine learning model.

Abstract: Methods and systems for operating axles of a vehicle are provided. In one example, a propulsion source of a first axle is operated in a torque control mode at a first torque and a propulsion source of a second axle is operated in a torque control mode at a second torque. Torque of the propulsion sources may be adjusted as a function of steering angle.

Abstract: This disclosure relates to an electrified vehicle and a method for gradually adjusting a displayed state of charge. An exemplary electrified vehicle includes a battery, a display configured to display a state of charge of the battery, and a controller configured to adjust the displayed state of charge such that the displayed state of charge gradually converges to an estimated state of charge of the battery.

Abstract: A computer, including a processor and a memory, the memory including instructions to be executed by the processor to determine a second convolutional neural network (CNN) training dataset by determining an underrepresented object configuration and an underrepresented noise factor corresponding to an object in a first CNN training dataset, generate one or more simulated images including the object corresponding to the underrepresented object configuration in the first CNN training dataset by inputting ground truth data corresponding to the object into a photorealistic rendering engine and generate one or more synthetic images including the object corresponding to the underrepresented noise factor in the first CNN training dataset by processing the simulated images with a generative adversarial network (GAN) to determine a second CNN training dataset. The instructions can include further instructions to train a CNN to using the first and the second CNN training datasets.

Abstract: A method of controlling a steering system on a marine vessel includes, in response to receiving a user input to engage a quick steer mode, employing a reduced steering ratio to translate positions of a steering wheel to desired steering angles of a marine drive. A vessel speed of a marine vessel is determined and then compared to a threshold vessel speed. An output limit is determined to prevent the marine vessel from further exceeding the threshold vessel speed while the quick steer mode is engaged. The marine drive is automatically controlled based on the output limit and a steering actuator associated with the marine drive is controlled based on the reduced steering ratio.

Abstract: Artificial-intelligence-based river information system. In an embodiment, a first training dataset is used to train a travel time prediction model to predict a travel time along the waterway for a given trip. In addition, a second training dataset is used to train a river level prediction model to predict a river level along the waterway for a given time. For each of a plurality of trips, a request is received that specifies the trip and a time of the trip, and, in response to the request, the travel time prediction model is used to predict a travel time for the trip, and the river level prediction model is used to predict a river level of the waterway at one or more points along the trip. Then, a voyage plan is generated based on one or both of the predicted travel time and the predicted river level.

Abstract: This document describes Kurtosis based pruning for sensor-fusion systems. Kurtosis based pruning minimizes a total quantity of comparisons performed when fusing together large sets of data. Multiple candidate radar tracks may possibly align with one of multiple candidate visual tracks. For each candidate vision track, a weight or other evidence of matching is assigned to each candidate radar track. An inverse of matching errors between each candidate vision and each candidate radar track contributes to this evidence, which may be normalized to produce, for each candidate vision track, a distribution associated with all candidate radar tracks. A Kurtosis or shape of this distribution is calculated. Based on the Kurtosis values, some candidate radar tracks are selected for matching and other remaining candidate radar tracks are pruned. The Kurtosis aids in determining how many candidates to retain and how many to prune.

Abstract: Pixel image data of a scene is received in which the pixel image data includes a two-dimensional representation of an object in the scene. Point cloud data including three-dimensional point coordinates of a physical object within the scene corresponding to the two-dimensional representation of the object is received. The three-dimensional point coordinates include depth information of the physical object. The point cloud data is mapped to an image plane of the pixel image data to form integrated pixel image data wherein one or more pixels of the pixel image data have depth information integrated therewith. A three-dimensional bounding box is predicted for the object using a convolutional neural network based upon the integrated pixel image data.

Abstract: Traversing a vehicle transportation network includes operating a scenario-specific operational control evaluation module instance. The scenario-specific operational control evaluation module instance includes an instance of a scenario-specific operational control evaluation model of a distinct vehicle operational scenario. Operating the scenario-specific operational control evaluation module instance includes identifying a multi-objective policy for the scenario-specific operational control evaluation model. The multi-objective policy may include a relationship between at least two objectives. Traversing the vehicle transportation network includes receiving a candidate vehicle control action associated with each of the at least two objectives. Traversing the vehicle transportation network includes selecting a vehicle control action based on a buffer value.

Abstract: Methods and devices for generating a trajectory for a self-driving car (SDC) are disclosed. The method includes: for a given one of the plurality of trajectory points of a given trajectory: (i) determining a presence of plurality of dynamic objects around the SDC, (ii) applying a first algorithm to determine a set of collision candidates, (iii) generating a segment line for the SDC, (iv) generating a bounding box for each of set of collision candidates, (v) for a given one of the set of collision candidates, determining a distance between the segment line and the respective bounding box to determine a separation distance, (vi) in response to the separation distance being lower than a threshold, determining that the given one of the set of collision candidates would cause the collision with the SDC, (vii) amending at least one of the plurality of trajectory points to render a revised candidate trajectory.

Abstract: A method, a system, and a computer program product may be provided for generating map data associated with one or more objects in a region. The method includes receiving point cloud data associated with the region. The method further includes generating a georeferenced top down raster image of the region, based on the point cloud data. The georeferenced top down raster image is indicative of a top surface of each of the one or more objects. The method further includes determining boundary data of the top surface of each of the one or more objects, based on the georeferenced top down raster image and generating the map data associated with the one or more objects, based on the boundary data of the top surface of each of the one or more objects.

Abstract: An apparatus and a method for providing traffic information are provided. The apparatus includes a traffic data database (DB) that stores traffic data and a processor connected to the traffic data DB. The processor generates a prediction model to predict a traffic flow and determines an appropriate data input range for a target time point in the future using the prediction model. Additionally, the processor extracts past traffic data from the traffic data DB based on the determined appropriate data input range, predicts a traffic flow at the target time point based on the extracted past traffic data and provides the predicted traffic flow as traffic information.

Abstract: A method, computer program product, and a system where a processor(s) determines generates a cognitive user profile representing patterns of usage of each of a plurality of users of the transportation resource sharing system, a cognitive resource profile for each resource of the plurality of resources, a cognitive route profile for each route traversed by at least one resource of the plurality of resources, and a cognitive station profile for each station of the plurality of stations. The processor(s) assigns one or more specific resources of the plurality of resources to one or more specific users of the plurality of users and the one or more specific resources of the plurality of resources to one or more specific stations of the plurality of stations.

Abstract: A method and apparatus for positioning a vehicle. The method may include: acquiring an identification element and non-identification elements of a current vehicle. The method can further include matching the identification element of the current vehicle with position elements in a high-precision map to determine an initial position of the current vehicle. The method can further include using at least one position element in the high-precision map to perform observational constraints on the non-identification elements of the current vehicle to acquire position elements corresponding to the non-identification elements of the current vehicle. The method can further include adjusting the initial position of the current vehicle using the position elements corresponding to the non-identification elements of the current vehicle to obtain a target position of the current vehicle.

Abstract: Embodiments of the present application provide a collision detection method and apparatus based on an autonomous vehicle, a device and a storage medium, where the method includes: acquiring first point cloud data of each obstacle in each region around the autonomous vehicle, where the first point cloud data represents coordinate information of the obstacle and the first point cloud data is based on a world coordinate system; converting the first point cloud data of the each obstacle into second point cloud data based on a relative coordinate system, where an origin of the relative coordinate system is a point on the autonomous vehicle; determining, according to the second point cloud data of the each obstacle in all regions, a possibility of collision of the autonomous vehicle. A de-positioning manner for collision detection is provided, thereby improving the reliability and stability of collision detection.