A vehicle control device includes a recognizer that is configured to recognize a surrounding situation of a host vehicle and a driving controller that is configured to control acceleration or deceleration and steering of the host vehicle on the basis of a recognition result of the recognizer. The driving controller is configured to cause the host vehicle to operate in at least any of a first driving state and a second driving state in which a rate of automation is higher or tasks required of an occupant are fewer than in the first driving state, and is configured to transition a driving state of the host vehicle to the first driving state on the basis of movement of a rearward vehicle that travels rearward of the host vehicle recognized by the recognizer in a direction of a vehicle width in a case where the host vehicle is operating in the second driving state.

The invention relates to a method for determining a lane change for a driver assistance system (100) of a vehicle (1), which method comprises calculating (220) a probability that other vehicles (2) are driving at a higher speed in a lane (L2) which is adjacent to a current lane (L1) of the vehicle (1), applying (240) a hysteresis to the calculated probability based on a driving parameter dependent on a last lane change, and issuing (250) a command to change lanes depending on the probability. The invention further relates to a driver assistance system and a vehicle which can carry out such a method.

A vehicle is provided to avoid a collision with a target object located in front of the vehicle by predicting an expected traveling path of the target object. The vehicle also predicts the possibility of a collision with the target object.

Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for generating unmapped U-turn predictions using a machine learning model. One of the methods includes obtaining features of an agent travelling on a roadway. One or more unmapped U-turn regions in a vicinity of the agent on the roadway are identified. For each of the unmapped U-turn regions and from at least the features of the agent, a respective likelihood score that represents a likelihood that the agent intends to make an unmapped U-turn at the unmapped U-turn region is generated. Based on the respective likelihood scores, one or more of the unmapped U-turn regions are selected. For each selected unmapped U-turn region, data specifying a candidate future trajectory in which the agent makes the unmapped U-turn at the selected unmapped U-turn region is provided as a possible future trajectory for the agent.

A communication system that determines a context and an intent of a specific remote vehicle located in a surrounding environment of a host vehicle includes one or more controllers for receiving sensed perception data including sensed perception data. The one or more controllers execute instructions to determine a plurality of vehicle parameters related to the specific remote vehicle. The the one or more controllers execute instructions to associate the specific remote vehicle with a specific lane of travel of a roadway based on map data. The one or more controllers determines possible maneuvers, possible egress lanes, and a speed limit for the specific remote vehicle for the specific lane of travel based on the map data, and determines the context and the intent of the specific remote vehicle based on the plurality of vehicle parameters, the possible maneuvers, the possible egress lanes for the specific remote vehicle, and the speed limit related to the specific remote vehicle.

A speed of a target vehicle in a target lane of operation is determined relative to a host vehicle in a host lane of operation. A virtual boundary is determined around the target vehicle based on the speed of the target vehicle. A position in the target lane and outside the virtual boundary is selected based on a) a first cost function for a deviation of a speed of the host vehicle from a requested speed, and b) a second cost function for a frequency of lane changes. Upon determining to move the host vehicle from the host lane to the target lane, the host vehicle is operated to the position in the target lane.

Techniques for determining unified futures of objects in an environment are discussed herein. Techniques may include determining a first feature associated with an object in an environment and a second feature associated with the environment and based on a position of the object in the environment, updating a graph neural network (GNN) to encode the first feature and second feature into a graph node representing the object and encode relative positions of additional objects in the environment into one or more edges attached to the node. The GNN may be decoded to determine a distribution of predicted positions for the object in the future. A predicted position of the object at a subsequent timestep may be determined by sampling from the distribution of predicted positions according to various sampling strategies. Alternatively, the predicted position of the object may be overwritten using a candidate position of the object.

A safe following distance estimation system and an estimation method thereof are provided. The safe following distance estimation system adapted for an autonomous vehicle includes a sensor and a processor. The sensor senses an adjacent vehicle to generate first sensing data, and senses the autonomous vehicle to generate second sensing data. The processor estimates a first friction parameter between wheels of the adjacent vehicle and a pavement according to pavement material data, and estimates a second friction parameter between wheels of the autonomous vehicle and the pavement according to the second sensing data. The processor calculates a safe following distance between the autonomous vehicle and the adjacent vehicle according to the first sensing data, the second sensing data, the first friction parameter, the second friction parameter.

Techniques for determining unified futures of objects in an environment are discussed herein. Techniques may include determining a first feature associated with an object in an environment and a second feature associated with the environment and based on a position of the object in the environment, updating a graph neural network (GNN) to encode the first feature and second feature into a graph node representing the object and encode relative positions of additional objects in the environment into one or more edges attached to the node. The GNN may be decoded to determine a distribution of predicted positions for the object in the future that meet a criterion, allowing for more efficient sampling. A predicted position of the object in the future may be determined by sampling from the distribution.

Techniques for determining unified futures of objects in an environment are discussed herein. Techniques may include determining a first feature associated with an object in an environment and a second feature associated with the environment and based on a position of the object in the environment, updating a graph neural network (GNN) to encode the first feature and second feature into a graph node representing the object and encode relative positions of additional objects in the environment into one or more edges attached to the node. The GNN may be decoded to determine a predicted position of the object at a subsequent timestep. Further, a predicted trajectory of the object may be determined using predicted positions of the object at various timesteps.