A location of a first object can be determined in an image. A line can be drawn on the first image based on the location of the first object. A deep neural network can be trained to determine a relative location between the first object in the image and a second object in the image based on the line. The deep neural network can be optimized by determining a fitness score that divides a number of deep neural network parameters by a performance score. The deep neural network can be output.

Among other things, we describe techniques for implementing a vehicle response to sensor failure. In general, one innovative aspect of the subject matter described in this specification can be embodied in methods that include receiving information from a plurality of sensors coupled to a vehicle, determining that a level of confidence of the received information from at least one sensor of a first subset of sensors of the plurality of sensors is less than a first threshold, comparing a number of sensors in the first subset of sensors to a second threshold, and adjusting the driving capability of the vehicle to rely on information received from a second subset of sensors of the plurality of sensors, wherein the second subset of sensors excludes the at least one sensor of the first subset of sensors.

A route risk mitigation system and method using real-time information to improve the safety of vehicles operating in semi-autonomous or autonomous modes. The method mitigates the risks associated with driving by assigning real-time risk values to road segments and then using those real-time risk values to select less risky travel routes, including less risky travel routes for vehicles engaged in autonomous driving over the travel routes. The route risk mitigation system may receive location information, real-time operation information, (and/or other information) and provide updated associated risk values. In an embodiment, separate risk values may be determined for vehicles engaged in autonomous driving over the road segment and vehicles engaged in manual driving over the road segment.

Systems and methods for characterizing a driving style of an autonomous vehicle are presented. A system may include one or more sensors configured to collect information concerning driving characteristics; a memory containing computer-readable instructions for evaluating the driving characteristics for a pattern(s) correlatable with a driving style of the autonomous vehicle and for characterizing aspects of driving style based on the one or more patterns; and a processor configured to evaluate the driving characteristics for the one or more patterns correlatable with the driving style, and characterize aspects of the driving style based on the pattern(s). Corresponding methods and non-transitory media are disclosed.

A vehicle includes: a plurality of sensor devices that determine a driver state; a driver state determining device that receives detection results from a plurality of sensor devices and determines whether the driver state is a dangerous state; and a driving assistance device that performs lane keeping control and speed control of a vehicle and transmits a network connection request to a management server when the driver state determining device has determined the dangerous state.

Systems, devices, computer-implemented methods, and/or computer program products that facilitate dynamic curve speed control adaptive to road conditions. In one example, a system can comprise a process that executes computer executable components stored in memory. The computer executable components can comprise a curvature component, a road condition component, and a safety component. The curvature component can generate composite curvature data for a curve of a road preceding a vehicle using digital map data and lane marker data. The road condition component can generate friction data for a surface of the road using sensor data obtained from an on-board sensor of the vehicle. The safety component can determine a safe operational profile for traversing the curve using the composite curvature data and the friction data.

An autonomous vehicle (AV) includes features that allows the AV to comply with applicable regulations and statues for performing safe driving operation. An example method for operating an AV includes receiving, from a sensor located on the AV, sensor data that captures a road sign located at a distance from the AV that is operating on a roadway; obtaining, from the sensor data, roadway information indicated by the road sign that corresponds to a segment of the roadway associated with the road sign that is ahead of a current position of the AV on the roadway; determining trajectory-related information for the AV for the distance that is based on the roadway information obtained from the sensor data; and causing the AV to travel in accordance with the trajectory-related information until a determination that the AV has arrived within the segment of the roadway associated with the road sign.

A method, apparatus, and system are described. The method includes generating a set of current values associated with at least one component included on a moving vehicle and providing the set of current values over a wireless network. The values are generated by one or more sensors. An edge computing device receives the current values. The method further includes processing the set of current values in real-time using at least one machine learning algorithm to identify a value of a point in time for a failure of one of the at least one component based on the set of current values and at least one set of past values received. The past values are stored in a memory. The set of current values are transmitted with a low time latency between the generating the set of current values and the processing of the set of current values.

An information processing apparatus acquires vehicle state information for each of a plurality of vehicles. The information processing apparatus estimates road surface state information of a road surface on which each of the vehicles has traveled, based on the acquired vehicle state information for each of the vehicles. The information processing apparatus estimates the road surface state information by inputting the acquired vehicle state information to a trained model that outputs the road surface state information in a case where the vehicle state information is input and that has been trained in advance based on training data in which the vehicle state information and the road surface state information are associated with each other.

The disclosed computer-implemented method may include monitoring, during a ride provided by an autonomous vehicle (AV), passenger communications between a passenger and a remote agent using an in-vehicle electronic device. The method may further include identifying passenger ride preferences based on a passenger request, and in association with a first occurrence of a ride event. The method may also include providing confirmation, via the in-vehicle electronic device, that the request is being fulfilled, the fulfillment of which causes a change in the passenger's AV experience based upon changes to features of the AV. The method may further include generating, based upon the request, a prediction of passenger ride preferences for the passenger and then, during a subsequent AV ride carrying the passenger, applying the predicted passenger ride preferences to an AV during a second occurrence of the ride event. Various other methods, systems, and computer-readable media are also disclosed.