NURO, INC.

A teleoperations system enables remote teleoperations monitoring and/or control of a vehicle. The teleoperations system may be validated and a teleoperations operator of the teleoperations system may be evaluated using interactive validation scene playbacks. Scenes may be recorded during the operations of an autonomous vehicle. The scenes may be programmatically analyzed to identify those that are suitable for validation playbacks and/or to determine one or more validation requirements of the scene. During an interactive validation scene playback, a scene may be retrieved from a database and the scene, including one or more video streams of the scene, may be presented to a teleoperations operator on a teleoperations human interface system. One or more validation inputs are received during the playback and the inputs are evaluated to determine whether they satisfy the validation requirements of the scene.

According to one aspect, methods to validate an autonomy system in conjunction with a teleoperations system are provided. A computing device obtains simulation data of an autonomous driving system of a vehicle. The simulation data includes first environment data representing an environment in which the vehicle is driven using the autonomous driving system. The computing device obtains teleoperations driving data of a teleoperations driving system by which the vehicle is remotely controlled with an aid of a human. The teleoperations driving data includes second environment data representing the environment in which the vehicle is driven using the teleoperations driving system. The computing device determines whether the autonomous driving system in conjunction with the teleoperations driving system meets a minimum deployment standard based on the simulation data and the teleoperations driving data.

According to one aspect, methods for simulating operations of an autonomous vehicle in which latencies associated between various components of the autonomous vehicle are injected into a simulation for validating an autonomous driving system (autonomy system and/or simulation system) for driving a vehicle are provided. A computing device obtains simulation data that represents an environment in which a vehicle is driven for a predetermined time using an autonomous driving system. The computing device determines at least one latency amount associated with operating the vehicle using the autonomous driving system, based on characteristics of the simulation data and injects the at least one latency amount into the simulation data, forming a simulation scenario. The computing device validates the autonomous driving system for driving the vehicle using the simulation scenario.

According to one aspect, a method includes obtaining an input and identifying at least a first vehicle which is to be commanded to perform at least one diagnostics test. The method also includes providing a remote diagnostics command to the first vehicle, the remote diagnostics command configured to command the first vehicle to perform at least a first diagnostics test, wherein the first diagnostics test is performed by the first vehicle on at least a first system included on the first vehicle. The first vehicle is monitored during the first diagnostics test. Diagnostics data associated with the first diagnostics test is obtained from the first vehicle and processed to generate diagnostics for the first vehicle.

According to one aspect, methods to validate an autonomy system using scenes of various complexities are provided. A computing device obtains a plurality of scenes. Each of the plurality of scenes represents an environment in which a vehicle is driven for a predetermined time. The computing device determines a complexity of each of the plurality of scenes based on characteristics of a respective scene and at least one maneuver required from an autonomous driving system to drive the vehicle through the respective scene without a collision. The computing device calculates a complexity score for each of the plurality of scenes based on the complexity determined for the respective scene and validates the autonomous driving system for driving the vehicle using a subset of the plurality of scenes selected based on the complexity score.

An autonomous vehicle can include multiple low voltage power domains. Each low voltage power domain can comprise a low voltage power distribution unit (LVPDU). An LVPDU can include a failsafe switch that can selectively couple or de-couple the low voltage power domain from a shared backup low voltage power source, which can be configured to concurrently supply backup power multiple low voltage power domains or LVPDUs. A first LVPDU of a first low voltage power domain can generate a control signal to control the failsafe switch of a second LVPDU of the second low voltage power domain to disconnect the second low voltage power domain from the shared low voltage backup power source, and vice versa.

According to one aspect, methods to select and implement a failover behavior based on situational awareness, when an autonomous vehicle enters a failover condition, are provided. Operation design domains, other users who share a road with the autonomous vehicle, and tailgaters may be considered when determining a deceleration profile, a trajectory length, and a trajectory type for the autonomous vehicle to follow when executing failover behavior. An autonomous driving system of an autonomous vehicle detects a failover event while the autonomous vehicle is traveling along a path and analyzes an environment surrounding the autonomous vehicle to determine whether a tailgater is present behind the autonomous vehicle. The autonomous driving system generates a failover plan for performing a failover behavior based on the failover event and the environment surrounding the autonomous vehicle and engages the autonomous vehicle to implement the failover plan for performing the failover behavior.

According to one aspect, an autonomous vehicle that includes a lidar unit collects lidar point cloud data that includes false returns or false positives, and characterizes the data associated with the false returns or false positives as drivable or not drivable. The false returns or false positives may be phantom points that are not associated with actual objects which may pose collision risks. Analyzing lidar point cloud data to characterize false returns or false positives as either drivable or not drivable enables an autonomous vehicle to operate efficiently by not having to avoid non-existent collision risks. False positives may be indicated when a wet or icy road surface acts as a mirror which reflects objects, and when precipitation such as raindrops appear as objects. Characterizing such false positives as drivable facilitates the efficient operation of an autonomous vehicle as the autonomous vehicle may drive over a mirror and/or through precipitation.

Collision filtering for autonomous vehicle simulations is disclosed herein. A simulation system can be configured to identify a simulated collision involving a simulated autonomous vehicle and a simulated agent representing an obstacle with which the simulated autonomous vehicle may collide. For example, the simulated collision can be simulated based on logged data from a real-world autonomous vehicle. The simulation system can classify the simulated collision as unrealistic or realistic. For example, classification can involve considering a nature of the simulated agent (including, e.g., whether the simulated agent is a drivable obstacle), a nature of the collision, whether the collision is avoidable through an evasive maneuver, whether the collision is similar to a known false positive, and/or input from a human operator. The simulation system can filter the simulated collision from participating in a collision triage analysis in response to classifying the simulated collision as an unrealistic simulated collision.

According to one aspect, data collected from various modules on an autonomous vehicle is synchronized, and snapshots of the synchronized data, referred to herein as ‘data snapshots’, are created. Contextual information associated with the data snapshots, e.g., indexes and/or metadata, is collected and/or derived, and the contextual information is stored along with the snapshots. The data snapshots may then be substantially searched, using the contextual information, to identify particular scenes for analysis. In addition to searching for particular scenes, series of scenes may be searched using the contextual information such that scenarios encountered by the autonomous vehicle over time may be identified.