One or more outputs from a planning module of an ADV are received. Data of a driving environment of the ADV is received. A performance of the planning module is evaluated by determining a score of the performance of the planning module based on the data of the driving environment and the one or more outputs from the planning module. Whether the one or more outputs from the planning module violates at least one of a set of safety rules is determined. The score is determined being larger than a predetermined threshold in response to determining that the one or more outputs from the planning module violate at least one of the set of safety rules. Otherwise, the score is determined based on a machine learning model. The planning module is modified by tuning a set of parameters of the planning module based on the score.

A method for setting a tuning parameter for an Automated Driving System (ADS) of a vehicle is disclosed. A corresponding non-transitory computer-readable storage medium, vehicle control device and a vehicle comprising such a control device are also disclosed. The method comprises receiving environmental data from a perception system of the vehicle, said environmental data comprising a plurality of environmental parameters, determining, by means of a self-learning model, an environmental scenario based on the received environmental data; setting the tuning parameter for the ADS based on the self-learning model and the determined environmental scenario, the tuning parameter defining a dynamic parameter of the ADS, receiving at least one signal representative of a vehicle user feedback on the set tuning parameter, and updating the self-learning model for the set tuning parameter for the identified environmental scenario based on the received vehicle user feedback.

A computer-implemented method of adaptively controlling an autonomous operation of a vehicle is provided. The method includes steps of (a) in a critic network in a computing system configured to autonomously control the vehicle, determining, using samples of passively collected data and a state cost, an estimated average cost, and an approximated cost-to-go function that produces a minimum value for a cost-to-go of the vehicle when applied by an actor network; and (b) in an actor network in the computing system and operatively coupled to the critic network, determining a control input to apply to the vehicle that produces the minimum value for the cost-to-go, wherein the actor network is configured to determine the control input by estimating a noise level using the average cost, a cost-to-go determined from the approximated cost-to-go function, a control dynamics for a current state of the vehicle, and the passively collected data.

Embodiments herein can determine an optimal route for an autonomous electric vehicle. The system may score viable routes between the start and end locations of a trip using a numeric or other scale that denotes how viable the route is for autonomy. The score is adjusted using a variety of factors where a learning process leverages both offline and online data. The scored routes are not based simply on the shortest distance between the start and end points but determine the best route based on the driving context for the vehicle and the user.

A vehicular arbitration system includes: a main manager configured to receive one or more requests from a plurality of first application execution units and to determine a request for operating a predetermined on-vehicle device based on the received one or more requests and a predetermined rule; and a plurality of sub-managers respectively configured to arbitrate the request determined by the main manager and a request input from at least one second application execution unit that is different from the plurality of first application execution units and to control the on-vehicle device based on an arbitration result.

Methods, systems, and apparatus, including computer programs encoded on computer storage media, for generating roadway crossing intent labels for training a machine learning model to perform roadway crossing intent predictions. One of the methods includes obtaining data identifying a training input, the training input including data characterizing an agent in an environment as of a given time, wherein the agent is located in a vicinity of a roadway in the environment at the given time. Future data characterizing (i) the agent, (ii) the environment or (iii) both over a future time period that is after the given time is obtained. From the future data, an intent label that indicates a likelihood that the agent intended to cross the roadway at the given time is determined. The training input is associated with the intent label in training data for training the machine learning model.

The present invention relates to methods and systems that utilize the production vehicles to develop new perception features related to new sensor hardware as well as new algorithms for existing sensors by using self-supervised continuous training. To achieve this the production vehicle's own perception output is fused with other sensors in order to generate a bird's eye view of the road scenario over time. The bird's eye view is synchronized with buffered sensor data that was recorded when the road scenario took place and subsequently used to train a new perception model to output the bird's eye view directly.

Embodiments of the present disclosure relate to generating velocity profiles for an autonomous vehicle (101). An ECU (107) of the autonomous vehicle (101) receives road information from one or more sensors (106) associated with the autonomous vehicle (101). One or more parameters related to smooth movement of the autonomous vehicle on the road is determined from the road information. Further, a first velocity profile is produced using an AI model and a second velocity profile is produced using a hierarchical model, based on the one or more parameters. Furthermore, one of the first and the second velocity profile is selected by comparing the first and the second velocity profiles. The selected velocity profile has a lower value of velocity value compared to the other velocity profile. The selected velocity profile is provided to the autonomous vehicle (101) for navigating on the road (102) smoothly.

A distracted driving analysis system for identifying distracted driving events is provided. The system includes a processor in communication with a memory device programmed to: (i) receive driving event records, each driving event record including phone usage by a user, wherein a driving event record is labeled as an actual distracted driving event or a passenger event, (ii) divide the driving event records into at least two clusters based at least in part upon common features and the labels of each driving event record by processing the plurality of driving event records with a semi-supervised machine learning algorithm, (iii) generate a trained model based at least in part upon the at least two clusters, (iv) process a new driving event using the trained model, (v) assign the new driving event to one of the clusters using the trained model, and/or (vi) determine whether the new driving event is an actual distracted driving event or a passenger event.

A vehicle control device includes: a target traveling path setting unit that sets a target traveling path of an own vehicle; a reference position setting unit that sets a reference position of the own vehicle for specifying a position of the own vehicle with respect to the target traveling path; and a control unit that controls a steering assist amount of a steering wheel, based on a positional deviation being a deviation between the target traveling path set by the target traveling path setting unit and the reference position of the own vehicle set by the reference position setting unit. The reference position setting unit changes the reference position according to a vehicle speed.