Abstract: Systems and methods for operating an autonomous vehicle (AV) are provided. The method includes detecting one or more objects in an environment, predicting a first set of predicted object trajectories comprising one or more trajectories for each of the detected one or more objects, generating a plurality of candidate AV trajectories for the AV, scoring each of the candidate AV trajectories according to a cost function, using the scoring to select a final AV trajectory for execution, determining which of the predicted object trajectories affected the final AV trajectory and which did not do so, adding the predicted object trajectories that affected the final AV trajectory to a persisted prediction cache, excluding from the persisted prediction cache any predicted object trajectories that did not affect the final AV trajectory, and executing the final AV trajectory to cause the AV to move along the final AV trajectory.

Abstract: A system includes a computer-readable memory for storage and retrieval of LiDAR data. The memory includes one or more point cloud data structures, each point cloud data structure including a first header block, and a plurality of point blocks. Each of the plurality of point blocks is configured to store at least a portion of LiDAR point cloud data. The memory includes one or more image data structures having a second header block, and a plurality of image column blocks. Each image column block includes a column of pixels of a corresponding image from the LiDAR point cloud data, and the image column blocks collectively contain all pixels of the corresponding image.

Abstract: Disclosed herein are embodiments for a control module cooling system. For example, the control module is provided with a chassis and a circuit board assembly that is mounted to the chassis and includes electronics that generate heat during operation. A cold plate is mounted to the chassis and adjacent to the circuit board assembly and defines an opening. A cold block is coupled to the cold plate and comprises: a base disposed within the opening and coupled to the electronics; a plurality of fins extending from the base and spaced apart from each other to form channels; and a plate that is spaced apart from the base to form a cavity. The plate defines an inlet that is arranged over a central portion of the plurality of fins to receive a liquid therethrough to facilitate impinging flow of the liquid through the channels to transfer heat from the electronics.

Abstract: Disclosed herein are system, method, and computer program product embodiments for handling unmapped speed limit signs. For example, the method includes receiving sensor data from a sensor of a vehicle, identifying a traffic sign within a field of view of the vehicle based on the sensor data, identifying one or more lane segments of a road associated with the identified traffic sign based on a determination that the identified traffic sign is not mapped in a priori map, and updating an attribute associated with a segment of the one or more lane segments of the road based on a determination that a detected attribute corresponding to the identified sign is more restrictive than an attribute associated with the lane segment in the a priori map.

Abstract: Methods of identifying corner case simulation scenarios that are used to train an autonomous vehicle motion planning model are disclosed. A system selects a scene that includes data captured by one or more vehicles over a time period. The data includes one or more actors that the vehicle's sensors perceived over the time period in a real-world environment. The system selects a scene that includes a safety threshold violation, and it identifies the trajectory of an actor that participated in the violation. The system generates simulated scenes that alter the trajectory of the actor in the selected scene, selects simulated scenes that are more likely to occur in the real world and that may include safety threshold violations that go beyond any that may be found in the original scene, and uses the selected simulated scenes to train an autonomous vehicle motion planning model.

Abstract: Systems, methods, and computer-readable media are disclosed for a systems and methods for improved smart infrastructure data transfer. An example method may involve identifying that a software update is available for a smart infrastructure system. The example method may also involve determining, by a processor of the smart infrastructure system and using a signal strength between a first vehicle and the smart infrastructure system, that the first vehicle is within a threshold range of the smart infrastructure system. The example method may also involve establishing, by the smart infrastructure system, a first ad-hoc peer-to-peer communication link with the first vehicle. The example method may also involve sending, to the vehicle, a request for the software update. The example method may also involve receiving, from the vehicle, at least a first portion of the software update that is transferred using the first ad-hoc peer-to-peer communication link.

Abstract: Disclosed herein are system, method, and computer program product embodiments for determining a precipitation rate. In some embodiments, a processor receives precipitation data from an array of sensors of a set of array of sensors disposed on a vehicle. A set of lidar sensors are disposed on the vehicle and the array of sensors is positioned within a predetermined distance of a respective lidar sensor of the set of lidar sensors. The processor generates a precipitation rate based on an average of the precipitation data received from the array of sensors and trains a model to control a function of the vehicle during precipitation using the precipitation rate.

Abstract: Disclosed herein are system, method, and computer program product embodiments for handling changes in road markings. For example, the method includes identifying a road marking in a road trajectory of a vehicle using an artificial intelligence (AI) model and sensor data from a sensor of the vehicle and performing a map update based on a determination that there is a change in road markings in the road trajectory. The determination is based on at least on-board data generated when the vehicle is traversing the road trajectory and off-board data generated using stored data. The sensor data includes two or more sensor modalities.

Abstract: Systems, methods, and computer-readable media are disclosed for obtaining data from multiple internal vehicle networks. An example method may include receiving, using a first internal communication network of a vehicle, a first query for data from one or more devices of the vehicle, wherein the first query is formatted for communication over the first internal communication network. The example method may also include generating, based on the first query, a second query associated with a supplemental communication network, wherein the supplemental communication network is also internal to the vehicle. The example method may also include sending, using the supplemental communication network, the second query to a first device, the first device being located on the supplemental communication network. The example method may also include receiving, from the first device, data relating to the second query.

Abstract: Disclosed herein are system and method embodiments to implement a validation of an SfM map. An embodiment operates by receiving a motion-generated map corresponding to a digital image, generating a first depth map, wherein the first depth map comprises depth information for one or more triangulated points located within the motion generated image. The embodiment further receives a light detection and ranging (lidar) generated point cloud including at least a portion of the one or more triangulated points, splats the lidar point cloud proximate to the portion of the one or more triangulated points and generates a second depth map for the portion and identifies an incorrect triangulated point, of the one or more triangulated points, based on comparing the first depth information to the second depth information. The incorrect triangulated points may be removed from the SfM map or marked with a low degree of confidence.

Abstract: Disclosed herein are system, method, and computer program product aspects for enabling providing map updates to an autonomous vehicle (AV). The system can has four features that can (1) provide live map updates, (2) provide a mechanism to optimize map updates, (3) utilize AV sensors to provide local map updates and integrate into the broader map update system, and (4) provide a mechanism to provide low latency updates.

Abstract: Disclosed herein are system, method, and computer program product aspects for simulating map update functions in an autonomous vehicle (AV) system. The system can create a simulation environment to test the AV system using data collected during regular operations of the AV system. The data comprises (1) map updating documents from a cloud service, (2) local map updating documents, and (3) observation data. A map-deviation detection mechanism is used to process the observation data.

Abstract: Disclosed herein are system, method, and computer program product aspects for enabling providing map updates to an autonomous vehicle (AV). The system can has four features that can (1) provide live map updates, (2) provide a mechanism to optimize map updates, (3) utilize AV sensors to provide local map updates and integrate into the broader map update system, and (4) provide a mechanism to provide low latency updates.

Abstract: Disclosed herein are system, method, and computer program product aspects for enabling providing map updates to an autonomous vehicle (AV). The system can has four features that can (1) provide live map updates, (2) provide a mechanism to optimize map updates, (3) utilize AV sensors to provide local map updates and integrate into the broader map update system, and (4) provide a mechanism to provide low latency updates.

Abstract: Disclosed herein are system, method, and computer program product aspects for enabling providing map updates to an autonomous vehicle (AV). The system can has four features that can (1) provide live map updates, (2) provide a mechanism to optimize map updates, (3) utilize AV sensors to provide local map updates and integrate into the broader map update system, and (4) provide a mechanism to provide low latency updates.

Abstract: Devices, systems, and methods are provided for testing and validation of a cleaning system for a sensor such as a camera. A device may capture a first image of a target using the sensor, wherein the sensor is in a clean state, and wherein the target is in a line of sight of the sensor. The device may apply an obstruction to a portion of a lens of the sensor. The device may apply a cleaning system to the lens. The device may capture a post-clean image after applying the cleaning system. The device may determine a post-clean image quality score based on comparing the post clean image to the first image. The device may compare the post-clean image quality score to a validation threshold. The device may determine a validation state of the cleaning system based on the comparison.

Abstract: Devices, systems, and methods are provided for sensor health and regression testing. A sensor testing system may include a first plurality of sensor testing targets positioned on a first side of a vehicle at a first distance from the vehicle, a second plurality of sensor testing targets positioned on a second side of the vehicle at the first distance, and a first sensor testing target positioned at a second distance from the vehicle, the second distance further from the vehicle than the first distance. The first and second pluralities of sensor testing targets both may include three or more sensor testing targets for testing cameras and light detection and ranging (LIDAR) sensors of the vehicle. The first sensor testing target may be used to test at least one of camera or LIDAR data. The sensor testing system may identify degradation of sensor performance, triggering further analysis and/or repairs.

Abstract: Disclosed herein are system, method, and computer program product embodiments for a mobile offload station for disconnected terminal operation. For example, the method includes deploying a mobile offload station to within a predetermined proximity of an autonomous vehicle (AV) responsive to receipt of an instruction to ingest log data stored on a data storage device onboard the AV. The method further includes establishing, by the mobile offload station, a data communication link with the AV responsive to entry into the predetermined proximity. The method further includes storing, by the mobile offload station, on a storage device onboard the mobile offload station, the log data transferred via the data communication link from the data storage device of the AV.

Abstract: Disclosed herein are system, method, and computer program product embodiments for a mobile offload station for disconnected terminal operation. For example, the method includes deploying at least one autonomous vehicle (AV) to within a predetermined proximity of a mobile offload station responsive to log data stored on a data storage device onboard the AV exceeding a data storage threshold. The method further includes establishing, by the at least one AV, a data communication link with the mobile offload station responsive to entry into the predetermined proximity. The method further includes transmitting, by the at least one AV, to a storage device onboard the mobile offload station, at least a portion of the log data via the data communication link.

Abstract: Devices, methods, and systems may obtain at least one point cloud, segment points in the at least one point cloud into a plurality of segments, train a neural network using known segments and a first loss function to generate a first trained neural network, train the first trained neural network using outlier segments and a second loss function to generate a second trained neural network, and train an extended isolation forest by applying an extended isolation algorithm to features of the known segments and features of the outlier segments to generate an anomaly score for each segment.