Abstract: A server includes a controller programmed to obtain information about a first perception model installed in a vehicle. The controller is further programmed to determine a value of updating the first perception model. The controller is further programmed to determine whether the first perception model needs to be updated to a second perception model based on the value of updating the first perception model. The controller is further programmed to transmit the second perception model to the vehicle in response to determining that the first perception model needs to be updated to the second perception model.

Abstract: A method comprises receiving an image of a road captured by a vehicle driving on the road, receiving a map of the road, the map comprising a road geometry of the road, obtaining an edge map of the road based on the image of the road, inputting the image, the map of the road, and the edge map into a trained regressor neural network, determining an estimated snow depth for each of one or more lanes of the road based on an output of the regressor neural network, and transmitting the estimated snow depth to an edge computing device.

Abstract: A method comprises receiving, at a vehicle system of a vehicle, sensor data from one or more sensors; detecting an object based on the sensor data; determining whether a dynamic map maintained by a remote computing device includes the detected object; upon determination that the dynamic map includes the detected object, determining a remaining time-to-live associated with the detected object based on data associated with the dynamic map; and transmitting data about the detected object to the remote computing device if the determined remaining time-to-live associated with the detected object is less than a predetermined threshold time.

Abstract: A vehicle for interpreting a traffic scene is provided. The vehicle includes one or more sensors configured to capture an image of an external view of the vehicle, and a controller. The controller is configured to obtain the captured image of the external view of the vehicle from the one or more sensors, segment a plurality of instances from the captured image, determine relational information among the plurality of instances, and generate a hyper graph including a plurality of nodes representing the plurality of instances and a plurality of edges representing the relational information among the plurality of instances.

Abstract: Candidate routes, from an origination to a destination, can be produced. Information can be obtained. The information can be of likelihoods of encounters, along the candidate routes, with types of driving situations demonstrated to cause changes in a degree of confidence, of an occupant of an autonomous vehicle, that a controller of the autonomous vehicle will be able to control the autonomous vehicle so that a result of the encounters is not a collision. Take-over probabilities can be obtained. The take-over probabilities can be of likelihoods that the occupant, in response to the encounters, will cause control of the autonomous vehicle to be transferred to the occupant. Based on the information and take-over probabilities, a route from the origination to the destination can be selected from the candidate routes. The autonomous vehicle can be caused to be configured to commence to move in a direction in accordance with the route.

Abstract: In accordance with one embodiment of the present disclosure, a system for generating surrounding views includes a controller. The controller may be programmed to perform operations including receiving a status information from a remote source, receiving a surround view system (SVS) request having a set of requirements from a vehicle, identifying one or more remote sources based on the status information of the one or more remote sources and the set of requirements of the SVS request, and directing the one or more remote sources to stream SVS data to the vehicle.

Abstract: System, methods, and embodiments described herein relate to dynamically generating a wide field-of-view three-dimensional pseudo point cloud of an environment around a vehicle. A disclosed method may include capturing, via a camera, a first view in a first image, determining a first depth map based on the first image, obtaining, from an external system, a second image of a second view that overlaps the first view and a second depth map based on the second image, inputting the first image and second image into a self-supervised homograph network that is trained to output a homographic transformation matrix between the first image and the second image, and generating a three-dimensional pseudo point cloud that combines the first depth map and the second depth map based on the homographic transformation matrix.

Abstract: Connected vehicles and methods described herein provide for message sender identification. Connected vehicles and methods disclosed herein can include generating a message based hyper-graph based on the positions of connected elements of a road traffic network as communicated to the connected vehicle by the connected elements. Connected vehicles and methods disclosed herein can include generating, a perception based hyper-graph based on the perceived positions of at least a subset of the elements of the road traffic network as determined by a sensor of the vehicle. Connected vehicles and methods disclosed herein can include matching a node of the message-based hyper-graph to a corresponding node of the perception-based hyper-graph to determine the sender of a received message.

Abstract: Methods, vehicles, and systems described herein generate alerts based on a generated macro state level indicative of a traffic hazard risk in view of a set of parameters which may lead driver to make various sub-conscious decisions. The set of parameters can include at least one of a vehicle parameter, driver state parameter, or external parameter.

Abstract: In one embodiment, a vehicle merge control system generates sensor data that indicate a detected vehicle in an environment of the ego vehicle, identifies a conflict based at least in part on the sensor data indicating that the detected vehicle and the ego vehicle are traveling: 1) in adjacent lanes that will merge into a single lane, and 2) at respective speeds that will result in the detected vehicle entering within a threshold range of the ego vehicle, determines, upon identification of the conflict, merge position assignments for the ego vehicle and the detected vehicle based at least in part on a game theory cost function applied to game theory actions that result in the merge position assignments exclusively being either a lead vehicle or a follower vehicle, and outputs an acceleration rate to achieve the merge position assignment for the ego vehicle.

Abstract: Driver classification systems and methods are disclosed herein. The driver classification method includes collecting first vehicle driving data from a first vehicle, processing the first vehicle driving data using a driver classification learning model including a machine learning algorithm at one of an edge server and the first vehicle to assign a driver classification to the first vehicle, updating the driver classification learning model based on additional driver classification learning models received from a plurality of additional vehicles, sending the driver classification to an insurance provider, receiving an insurance rate for the first vehicle from the insurance provider based on the driver classification of the first vehicle, and providing the insurance rate to the first vehicle.

Abstract: Described are systems and methods for substituting missing graph information caused during the transmission of graph information from one entity to another. In one example, a system includes a processor and a memory with machine-readable instructions that cause the processor to receive an incoming graph stream comprising a plurality of graphs having detected object information and at least one missing graph, transform the plurality of graphs into incoming embedded vectors having values that represent detected object information of the incoming graph stream, provide the embedded vector that corresponds to the at least one missing graph with a missing graph value, and substitute the missing graph value with a substitute value using a recurrent neural network that interpolates the substitute value using the incoming embedded vectors that are not equal to the missing graph value.

Abstract: Systems and methods for protecting a vehicle at an intersection are disclosed herein. One embodiment detects that the vehicle is stopped at the intersection at one of a first position and a second position; detects that a driver of the vehicle is pressing on an accelerator pedal of the vehicle; delays acceleration of the vehicle automatically by a first predetermined period, when the vehicle has been detected at the first position; and delays acceleration of the vehicle automatically by a second predetermined period, when the vehicle has been detected at the second position. Delaying acceleration of the vehicle automatically prevents the vehicle from being struck by a cross-traffic vehicle that has proceeded through the intersection against a first traffic signal in a red-light state.

Abstract: Systems and methods for maintaining a service session in an automotive edge computing environment are disclosed herein. One embodiment computes a first estimation result pertaining to computation resources of a current edge server with which a vehicle is in communication; computes a second estimation result pertaining to link quality; computes a third estimation result pertaining to service quality; compares the first, second, and third estimation results with first, second, and third predetermined thresholds, respectively; selects a new edge server with which to maintain the service session and initiates a service migration request, when the first predetermined threshold is triggered; initiates a wireless access handover request, when only the second predetermined threshold is triggered; and selects a new edge server with which to maintain the service session and initiates a service migration request, when only the third predetermined threshold is triggered and one of two predetermined conditions is met.

Abstract: System, methods, and other embodiments described herein relate to improving monitoring of traffic flows. In one embodiment, a method includes aggregating perception data associated with a road network from information sources to a server over a network. The method also includes generating a graph structure from the perception data in association with a neural network model. The graph structure is an incomplete representation of the road network in view of missing data. The method also includes completing the graph structure using the neural network model that forms a graph model of the traffic flows to de-noise the graph structure according to road constraints between two points in the road network. The method also includes communicating the graph model of the traffic flows to a vehicle to navigate traffic in the road network.

Abstract: Decentralized ridesharing systems and methods for assigning a user request to a particular registered vehicle. The method includes organizing a geographic area into a plurality of geographic zones, each of the plurality of geographic zones serviced by a corresponding regional server, each regional server monitoring a location and a passenger status of one or more registered vehicles within a corresponding geographic zone, and allocating dispatching and ridesharing tasks between the regional servers and the one or more registered vehicles within each of the geographic zones, wherein the regional servers and the one or more registered vehicles are each equipped with decision-making modules powered by a graph neural network with reinforcement learning.

Abstract: Systems and methods for personalizing adaptive cruise control in a vehicle are disclosed herein. One embodiment collects vehicle-following-behavior data associated with a particular driver; trains a Gaussian Process (GP) Regression model using the collected vehicle-following-behavior data to produce a set of adaptive-cruise-control (ACC) parameters pertaining to the particular driver, the set of ACC parameters modeling learned vehicle-following behavior of the particular driver; generates an acceleration command for the vehicle based, at least in part, on the set of ACC parameters; applies a predictive safety filter to the acceleration command to produce a certified acceleration command that has been vetted for safety; and controls acceleration of the vehicle automatically in accordance with the certified acceleration command to regulate a following distance between a lead vehicle and the vehicle in accordance with the learned vehicle-following behavior of the particular driver.

Abstract: A hybrid deterministic override to cloud based probabilistic advanced driver assistance systems. Under default driving conditions, an ego vehicle is controlled by a probabilistic controller in a cloud. An overall gap between the ego vehicle and a leading vehicle is divided into an emergency collision gap and a driver specified gap. The vehicle sensors monitor the overall gap. When the gap between the ego vehicle and the leading vehicle is less than or equal to the emergency collision gap, a deterministic controller of the ego vehicle overrides the cloud based probabilistic controller to control the braking and acceleration of the ego vehicle.

Abstract: A vehicle includes a controller programed to: collect a set of data related to a driver of the vehicle; predict a driving setting for the driver using the set of data and an initial student-T process (STP) machine learning (ML) model; generate an updated STP ML model based on the prediction of the driving setting as to the set of vehicle data; transmit incremental learning related to the updated STP ML model to a server; and receive, from the server, a personalized driving setting for the driver output from a cloud STP ML model trained by the incremental learning.

Abstract: Systems, methods, and other embodiments described herein relate to generating a notification about a risk of a target vehicle toppling over due to wind. In one embodiment, a method includes determining a wind force of the wind at a location of the target vehicle and determining one or more characteristics of the target vehicle. The method includes determining whether there is a risk of the target vehicle toppling over based on the wind force of the wind and the one or more characteristics of the target vehicle. The method includes generating a notification about the risk.