A radar sensor that uses electromagnetic band gap structures on a waveguide antenna surface is described. The radar sensor comprises a monolithic microwave integrated circuit (MMIC) and a printed circuit board (PCB) coupled to the MMIC. The radar sensor also comprises a waveguide antenna coupled to the PCB and having waveguide structures extending through the waveguide antenna to a surface of the waveguide antenna, the surface facing the PCB. The waveguide antenna also comprises double post electromagnetic band gap (EBG) structures disposed adjacent to and flush with broad sides of the waveguide structures on the surface of the waveguide antenna.
Systems and techniques are provided for determining error event likelihoods of routes of an autonomous vehicle (AV). An example method can determine, for each route from a plurality of routes, a respective set of route paths for the route, each route located within a geofence; determine, for each route path of each route, a first respective likelihood that the AV will use the route path for the route; determine, for each route path of each route, a second respective likelihood that the AV will encounter an error event along the route path during an autonomous operation; determine, for each route path of each route, an overall error event likelihood based on the first and second respective likelihoods for each route path of that route; and determine an aggregated error event likelihood for each route based on the overall error event likelihood of each route path of that route.
B60W 60/00 - Systèmes d’aide à la conduite spécialement adaptés aux véhicules routiers autonomes
B60W 50/02 - Détails des systèmes d'aide à la conduite des véhicules routiers qui ne sont pas liés à la commande d'un sous-ensemble particulier pour préserver la sécurité en cas de défaillance du système d'aide à la conduite, p. ex. en diagnostiquant ou en palliant à un dysfonctionnement
B60W 50/00 - Détails des systèmes d'aide à la conduite des véhicules routiers qui ne sont pas liés à la commande d'un sous-ensemble particulier
A method includes forming a temperature measurement assembly comprising a plurality of thermally coupled layers of a first thermally conductive material, wherein an inner layer of the plurality of thermally coupled layers comprises one or more voids defined therein, wherein each void receives a thermistor secured in the void using a second thermally conductive material that thermally couples the thermistor to the inner layer. The method also includes positioning the temperature measurement assembly on and in thermal communication with a surface of a device, and measuring temperatures of the surface of the device using the thermistors.
G01K 7/22 - Mesure de la température basée sur l'utilisation d'éléments électriques ou magnétiques directement sensibles à la chaleur utilisant des éléments résistifs l'élément étant une résistance non linéaire, p. ex. une thermistance
G01K 3/14 - Thermomètres donnant une indication autre que la valeur instantanée de la température fournissant des différences de valeursThermomètres donnant une indication autre que la valeur instantanée de la température fournissant des valeurs différenciées par rapport à l'espace
H01C 7/00 - Résistances fixes constituées par une ou plusieurs couches ou revêtementsRésistances fixes constituées de matériaux conducteurs en poudre ou de matériaux semi-conducteurs en poudre avec ou sans matériaux isolants
H01C 17/00 - Appareils ou procédés spécialement adaptés à la fabrication de résistances
4.
SYSTEM AND METHOD OF CAUSAL COMPOSITION DIFFUSION FOR CLOSED LOOP TRAFFIC GENERATION
A method includes receiving initial conditions for a traffic scenario including a plurality of interacting agents, the initial conditions defining states of the plurality of agents, identifying a causal structure among the plurality of agents based on the states of the plurality of agents, and ranking the plurality of agents based on the identified causal structure to determine a subset of key agents being most influential with respect to a controllability objective. For each agent of the plurality of agents, the method includes generating a future trajectory using a reverse sampling process of a diffusion model, and guiding the reverse sampling process by selectively applying a gradient of the controllability objective only to the determined subset of key agents.
G06F 30/27 - Optimisation, vérification ou simulation de l’objet conçu utilisant l’apprentissage automatique, p. ex. l’intelligence artificielle, les réseaux neuronaux, les machines à support de vecteur [MSV] ou l’apprentissage d’un modèle
A method for generating a motion plan for an autonomous agent includes obtaining scene data for an agent's environment, including map data and historical state data for one or more objects. A transformer-based encoder generates a set of scene embedding tokens from the scene data to represent a fixed environmental context. A transformer-based decoder autoregressively generates a sequence of action tokens representing a future trajectory. The generation of each subsequent action token is based on the scene embedding tokens and previously generated tokens. Each action token is selected from a discrete action space of unique Verlet actions, which represent accelerations. A future trajectory is determined from the sequence of action tokens and provided to a motion planning module of the autonomous agent.
A method adapts a pre-trained neural network model having layers with pre-trained weight matrices. At least one layer is augmented with a plurality of parameter-efficient adaptation modules, each module associated with a respective learnable scoring parameter. The model is fine-tuned on a target dataset while the pre-trained weight matrices are maintained in a frozen state. The fine-tuning includes performing a forward pass where an indicator function selectively applies a weight update from each module based on its scoring parameter and a threshold. A total loss value is determined from a task-specific loss and a sparsity-inducing regularization term. Parameters of the adaptation modules and the scoring parameters are updated based on the total loss value. A final, fine-tuned model having a sparse subset of activated adaptation modules is provided for an inference task.
G06N 3/0985 - Optimisation d’hyperparamètresMeta-apprentissageApprendre à apprendre
B60W 50/00 - Détails des systèmes d'aide à la conduite des véhicules routiers qui ne sont pas liés à la commande d'un sous-ensemble particulier
B60W 50/04 - Détails des systèmes d'aide à la conduite des véhicules routiers qui ne sont pas liés à la commande d'un sous-ensemble particulier pour surveiller le fonctionnement du système d'aide à la conduite
B60W 50/06 - Détails des systèmes d'aide à la conduite des véhicules routiers qui ne sont pas liés à la commande d'un sous-ensemble particulier pour améliorer la réponse dynamique du système d'aide à la conduite, p. ex. pour améliorer la vitesse de régulation, ou éviter le dépassement de la consigne ou l'instabilité
7.
UNSUPERVISED JOINT TRAINING OF MULTIMODAL PERCEPTION MODELS USING CURVATURE-BASED SAMPLING
A method includes obtaining Light Detection and Ranging (LIDAR) feature embeddings and obtaining camera feature embeddings. The method includes generating fusion feature embeddings by fusing the LIDAR feature embeddings and the camera feature embeddings. The method includes determining sampling weights for a plurality of points in a three-dimensional (3D) scene. The method includes selecting a subset of the plurality of points based on the sampling weights. The method includes determining a rendering loss by performing differentiable rendering on the selected subset of points. The method includes determining a prototype learning loss by comparing the LIDAR feature embeddings and the camera feature embeddings to a set of learnable prototypes representing parts of the 3D scene in a shared feature space. The method includes jointly training a LIDAR encoder, a camera encoder, and a fusion encoder based on the rendering loss and the prototype learning loss.
G06T 17/00 - Modélisation tridimensionnelle [3D] pour infographie
B60W 10/04 - Commande conjuguée de sous-ensembles de véhicule, de fonction ou de type différents comprenant la commande des ensembles de propulsion
B60W 10/18 - Commande conjuguée de sous-ensembles de véhicule, de fonction ou de type différents comprenant la commande des systèmes de freinage
B60W 10/20 - Commande conjuguée de sous-ensembles de véhicule, de fonction ou de type différents comprenant la commande des systèmes de direction
B60W 60/00 - Systèmes d’aide à la conduite spécialement adaptés aux véhicules routiers autonomes
G06T 7/521 - Récupération de la profondeur ou de la forme à partir de la télémétrie laser, p. ex. par interférométrieRécupération de la profondeur ou de la forme à partir de la projection de lumière structurée
G06T 7/55 - Récupération de la profondeur ou de la forme à partir de plusieurs images
G06T 7/64 - Analyse des attributs géométriques de la convexité ou de la concavité
8.
ITERATIVE ATTENTION SCOPING ON SPATIALLY CONTINGUOUS POINT CLOUD BUCKETS
A method includes receiving a three-dimensional (3D) point cloud and generating a plurality of buckets by applying a spatial hash function to coordinates of points in the 3D point cloud. The method includes arranging, based on the spatial hash function, the points into a contiguous block of memory. The method includes performing a plurality of attention iterations. Each attention iteration includes selecting a subset of the plurality of buckets to define an attention scope for the attention iteration, loading point features corresponding to the points in the attention scope from a contiguous block of memory into a cache, generating updated point features by performing a multi-head self-attention operation on the loaded point features. The method includes generating, based on the updated point features from the plurality of attention iterations, a feature representation of the 3D point cloud.
G06F 30/12 - CAO géométrique caractérisée par des moyens d’entrée spécialement adaptés à la CAO, p. ex. interfaces utilisateur graphiques [UIG] spécialement adaptées à la CAO
Systems and techniques are provided for crowdsourcing image data depicting autonomous vehicles (AV) to determine user feedback about the AVs based on the image data. An example method can include obtaining a frame captured by a client device and shared by a user of the client device, wherein the frame depicts an AV in a scene where the frame was captured; identifying the AV depicted in the frame based at least in part on a content of the frame; determining, based at least in part on the frame, an issue experienced by the AV in the scene; and determining an action to take in response to at least one of the frame shared by the user and the issue experienced by the AV.
Systems and techniques are disclosed for processing sensor data using neural networks. An example method can include receiving, via a first sensor, a first set of sensor data, wherein the first set of sensor data represents an environment; receiving, via a second sensor, a second set of sensor data, wherein the second set of sensor data represents the environment; generating one or more multi-view projection (MVP) tokens based on a combination of the first set of sensor data and the second set of sensor data, wherein each MVP token of the one or more MVP tokens includes a representation of the first set of sensor data projected into space and/or the second set of sensor data projected into space; and processing, via a neural network with an attention mechanism, the one or more MVP tokens.
Systems and methods for a multi-camera object detector having two or more backbone models. In particular, systems and methods are provided for including two or more backbone machine learning models, with one backbone optimized for speed and the other backbone optimized for precision. In particular, the first backbone can be highly accurate but slower than the second backbone and with a higher computer resource usage. The second backbone can be fast and efficient but have lower accuracy for object detection. In some examples, the second backbone can use fewer images and/or lower resolution images. The determination of which backbone to use can be based on fixed rules, or it can be determined based on another machine learning component. The outputs from the first and second backbones for each camera can be combined together into a unified representation, such as a bird's eye view (BEV) space.
G06V 10/70 - Dispositions pour la reconnaissance ou la compréhension d’images ou de vidéos utilisant la reconnaissance de formes ou l’apprentissage automatique
G06T 7/73 - Détermination de la position ou de l'orientation des objets ou des caméras utilisant des procédés basés sur les caractéristiques
G06V 10/77 - Traitement des caractéristiques d’images ou de vidéos dans les espaces de caractéristiquesDispositions pour la reconnaissance ou la compréhension d’images ou de vidéos utilisant la reconnaissance de formes ou l’apprentissage automatique utilisant l’intégration et la réduction de données, p. ex. analyse en composantes principales [PCA] ou analyse en composantes indépendantes [ ICA] ou cartes auto-organisatrices [SOM]Séparation aveugle de source
G06V 10/82 - Dispositions pour la reconnaissance ou la compréhension d’images ou de vidéos utilisant la reconnaissance de formes ou l’apprentissage automatique utilisant les réseaux neuronaux
12.
Systems and techniques for autonomous vehicle network selection based on passenger mobile device feedback
Systems and techniques are provided for an autonomous vehicle (AV) to select a mobile network. An example method can include establishing a connection between the AV and a mobile device associated with a passenger of the AV; identifying, using the connection, a first mobile network associated with the mobile device; determining that the first mobile network associated with the mobile device is different than a second mobile network associated with the AV; determining, based on a first network indicator associated with the first mobile network and a second network indicator associated with the second mobile network, that the first mobile network is preferable to the second mobile network; and initiating a mobile connection between the AV and the first mobile network.
H04W 76/11 - Attribution ou utilisation d'identifiants de connexion
H04W 4/40 - Services spécialement adaptés à des environnements, à des situations ou à des fins spécifiques pour les véhicules, p. ex. communication véhicule-piétons
H04W 76/15 - Établissement de connexions à liens multiples sans fil
Loss of connectivity between an AV with the back-office system may cause the vehicle to enter a degraded state, which may result in manual retrieval of the vehicle. The AV can maintain two redundant network data connections with two network operators using two dual-SIM, dual-standby network access devices. An algorithm can be implemented to select which two network operators to connect with using four different eSIMs. The algorithm can utilize locally measured network operator quality measurement data to make swapping decisions. The algorithm can utilize network operator connectivity quality information aggregated from network operator quality measurement data collected by many AVs.
H04W 36/30 - La resélection étant déclenchée par des paramètres spécifiques par des données de mesure ou d’estimation de la qualité des liaisons
H04W 36/32 - La resélection étant déclenchée par des paramètres spécifiques par des données de localisation ou de mobilité, p. ex. des données de vitesse
H04W 88/06 - Dispositifs terminaux adapté au fonctionnement dans des réseaux multiples, p. ex. terminaux multi-mode
A method includes projecting a first illumination pattern towards a scene. The first illumination pattern illuminates a first subset of a field of view. The method includes capturing a first measurement corresponding to light reflected from the scene in response to the first illumination pattern. The method includes projecting a second illumination pattern toward the scene. The second illumination pattern illuminates the second subset of the field of view. The method includes capturing a second measurement corresponding to light reflected from the scene in response to the second illumination pattern. The method includes comparing the first measurement with the second measurement and determining a depth value for an object within the scene based on comparing the first measurement with the second measurement. The method includes controlling movement of a vehicle based on the depth value for the object.
A process for mitigating potential disturbance can include obtaining at least one of wireless activity data associated with a geographic area, vehicle traffic data associated with the geographic area, and sound data associated with the geographic area; determining, based on one or more of the wireless activity data, the vehicle traffic data, and the sound data, a disturbance factor associated with the geographic area, wherein the disturbance factor indicates a likelihood of an autonomous vehicle activity disturbing one or more persons within the geographic area; and determining, based on the disturbance factor, instructions corresponding to at least one autonomous vehicle.
B60W 60/00 - Systèmes d’aide à la conduite spécialement adaptés aux véhicules routiers autonomes
B60W 40/04 - Calcul ou estimation des paramètres de fonctionnement pour les systèmes d'aide à la conduite de véhicules routiers qui ne sont pas liés à la commande d'un sous-ensemble particulier liés aux conditions ambiantes liés aux conditions de trafic
16.
Systems and Method for Tire Road Limit Nearness Estimation
For one embodiment of the present invention, a computer implemented method of determining tire road limit nearness estimation is described. The computer implemented method includes obtaining sensor signals from a sensor system of a vehicle to monitor driving operations and to determine localization of the vehicle. The method further includes determining lateral force disturbances for front and rear lateral accelerations and a bulk longitudinal force disturbance for the vehicle based on the localization and the sensor signals and determining a tire road limit nearness estimation for the vehicle based on the sensor signals, the lateral force disturbances for front and rear lateral accelerations and the bulk longitudinal force disturbance.
A printed circuit board assembly (PCBA) waveguide assembly comprises a PCBA, a waveguide, and a spring-loaded rivet. The spring-loaded rivet comprises a rivet head, a rivet shank, a rivet tail, and a spring element and fastens the PBCA to the waveguide. The spring element provides flexibility in the clamping force of the rivet to accommodate thermal expansion and contraction, creeping and shear forces experienced by the PCBA and waveguide.
A remote build orchestrator for building and testing a software program is described. The remote build orchestrator includes a generator service for generating a build graph from a definition of the software program input to the remote build orchestrator. The build graph includes a plurality of actions, wherein the actions are defined by inputs, outputs, and commands, and wherein outputs of actions that are dependent on other actions for inputs are represented in the build graph by placeholders. The remote build orchestrator also includes an unwinder service for receiving the build graph from the generator service and processing the build graph into a series of requests for execution of the actions. The remote build orchestrator further includes a remote build execution service for executing the actions in response to the received requests and returning results of the executing to the unwinder service.
A unified simulation interface system is described and includes an interface module for receiving a simulation build graph, wherein the simulation build graph comprises a static portion and a dynamic portion, wherein the interface module generates application programming interface (API) payloads for execution nodes comprising the static portion of the simulation build graph; a simulation generator module for generating API payloads for execution nodes comprising the dynamic portion of the simulation build graph; and a remote build execution service for receiving the API payloads generated by the interface module and the simulation generator module, and scheduling execution of tasks in connection with the received API payloads on a compute cluster.
Aspects of the subject technology relate to systems, methods, and computer-readable media for diversifying training data through application of a vision-language model. A subset of images can be separated from a plurality of images in a dataset based on a presence of a specific object associated with autonomous driving in the subset of images. The specific object can be segmented in a portion of the image in the subset of images through application of a vision-language model. Training data for training a model associated with AV operation can be augmented by inserting the portion of the image into the training data to generate augmented training data. The model can be trained with the augmented data.
Systems and methods for identifying potential collisions in which another vehicle may collide with the AV. The AV detects the other vehicle and determines the other vehicle speed, acceleration, and trajectory. The AV determines that there is a risk of collision if the other vehicle continues on its current trajectory at its current speed and its current acceleration. When an AV determines there is an imminent risk that another vehicle may collide with the AV, the AV can initiate a honk to alert the other vehicle of the potential collision. In some examples, the AV can identify a potential rear end collision, and the AV can honk at the vehicle approaching it from behind. In other examples, the approaching vehicle can come from any direction. Honking is designed to be limited so that an AV does not initiate too many honks.
A company that develops software applications for operating and managing autonomous vehicles has an array of developer teams working on many different projects. In addition, continuous integration and continuous delivery of software applications for the company can involve complex pipelines. A pipeline for an application can include parts, such as, integration testing, building, build testing, deploying, and monitoring the application. Different systems may be tasked to execute a part of the pipeline for the application on a destination (e.g., a particular namespace), and the respective systems may need access and/or permissions to execute the part of the pipeline on the destination. Adhering to the principle of least privilege access, permission(s) for executing the part of the pipeline may be narrowly scoped for the system executing the part, so that the system may have only what the system needs to execute the part of the pipeline, and no more.
G06F 21/57 - Certification ou préservation de plates-formes informatiques fiables, p. ex. démarrages ou arrêts sécurisés, suivis de version, contrôles de logiciel système, mises à jour sécurisées ou évaluation de vulnérabilité
Systems, methods, and non-transitory computer readable mediums are provided for testing and measuring the behavior or logic of one or more cloud services associated with a simulated autonomous vehicle (AV). For example, a system may be configured to receive log data of a simulated AV. In some examples, the log data may characterize one or more events that occurred during a simulation of the simulated AV. Additionally, the system may instantiate an AV bot based on the log data. Moreover, the system may generate, by the instantiated AV bot, a first set of messages. In some examples, the first set of messages may be based on the log data. Further, the system may provide the first set of messages to a cloud service.
Autonomous vehicles (AVs) may have processors to perform compute operations. These processors may be susceptible to overheating, due to one or more mechanical failures that may occur in the assembly that encloses the processors. To proactively address potential overheating of these processors, temperature of the processors can be monitored over time, e.g., at launch and during AV operation. A temperature profile line can be fitted to envelop the temperature data samples. The temperature profile line can indicate an extent of overheating, the rate of regression of the assembly's ability to cool the processors and predict when the processor may reach a critical temperature in the future. Based on one or more parameters that define the temperature profile line, it is possible to determine a regression profile of the processor. The regression profile can dictate whether to take an action to compensate for the regression.
A radar system comprises transmit antennas configured to transmit a radar signal and receive antennas that receive a radar return. The radar system further comprises circuitry configured to perform acts, the acts comprising generating a phase-shifted radar signal using a unique phase code that is commonly assigned to all transmitters included in the radar system. The acts further comprise transmitting the phase-shifted radar signal from transmit antennas and receiving a radar return at receive antennas. Additionally, the acts comprise, during signal processing, performing phase compensation by multiplying a radar signal received in the radar return by a phase compensation value that is commonly assigned to all receivers included in the radar system in order to remove the phase shift. The acts also comprise outputting an un-shifted processed radar signal.
A vehicle network logging system (VNLS) is described and includes a novel scenario detection and retention (NSD&R) module for receiving data segments comprising scenes, the NSD&R module configured to featurize the received segments based on the scenes comprising the segments, wherein the segments comprise sensor data collected by onboard sensors of a vehicle on which the VNLS is installed; assign the featurized segments to respective bins based on the featurization, wherein each of the bins is assigned to one of a plurality of coverage tiers; randomly select one of the segments from the bins assigned to a lowest one of the coverage tiers; and retain the selected one of the segments in the VNLS.
A sensor housing system comprises an upper housing portion, a lower housing portion, and a wire spring closure device having a closure feature configured to secure the upper housing portion to the lower housing portion. The wire spring closure device comprises an upper portion comprising an upper retaining bar, and a spring portion coupled to the upper portion. The wire spring closure device also comprises a lower portion coupled to the spring portion and comprising a lower retaining bar.
A company that develops software applications for operating and managing autonomous vehicles has an array of developer teams working on many different projects. Successful continuous delivery of software applications for the company can depend on reliability of the cluster infrastructure. Platform owners may be most knowledgeable when it comes to setting up cluster group that spans multiple regions. A developer working on a project can select, on a project management system, a cluster group for deployment of the project. A cluster group can specify a name for the cluster group, a plurality of environment-stack channels, and one or more clusters in deployment infrastructure associated with each environment-stack channel. The project management system can seamlessly provision resources for the project according to the selected cluster group.
A vehicle network logging system (VNLS) is described and includes a logger for storing data associated with events, wherein for each of the events, the associated data is created when a trigger associated with the event is received at the VNLS; and a compute node for, for each of the events, subsequent to occurrence of creation of data associated with the event, identifying a data tier for the event, the identified data tier selected from a plurality of data tiers of the VNLS and defined by the trigger associated with the event; and logging the data for the event to an internal disk of the VNLS in the identified data tier, wherein the identified data tier specifies an offload policy and a deletion policy for the data.
Vehicles detect temporary traffic restrictions and provide information describing the traffic restrictions to a remote computer system. The remote computer system can determine a routing cost for the traffic restriction and alert other vehicles in a fleet of vehicles about the traffic restrictions. The other vehicles can account for the routing cost when determining a route to follow; the route may avoid the traffic restriction. Vehicles in the fleet, when driving near a particular traffic restriction, may perceive the area where the traffic restriction was detected and provide updates about the traffic restriction, e.g., whether the boundaries of the TTR have changed (e.g., a construction area or emergency response has expanded or moved), or the TTR has been removed (e.g., a construction area has reopened, or a stopped vehicle has left the roadway).
Systems and techniques are provided for multi-stage training of a multi-network system. An example method can include training, using training data, a first neural network during a first training stage; generating, by the first neural network, one or more outputs; based on a determination that the first training stage and training of the first neural network are complete, providing the one or more outputs to a second neural network that has an input data dependency comprising data generated by the first neural network; and based on the determination that the first training stage and training of the first neural network are complete, training, using the one or more outputs from the first neural network, the second neural network during a second training stage initiated after completion of the first training stage.
The present disclosure generally relates to improved centroid predictions. In some aspects, a method of the disclosed technology includes: collecting, from a sensor, sensor data comprising data points; segmenting, via a first network, the sensor data into a first portion of the sensor data and a second portion of the sensor data; determining a semantic label for the first portion of the sensor data; determining local semantic information for each data point of the first portion of the sensor data; removing, using a point mask, the second portion of the sensor data; and determining, via a second network, a centroid of the first portion of the sensor data based on the first portion of the sensor data, the semantic label, and the local semantic information. Systems and machine-readable media are also provided.
The present disclosure generally relates to systems and techniques for mitigating electromagnetic interference and, more specifically, to using a filter printed circuit board (PCB) within an autonomous vehicle to mitigate electromagnetic interference. In some aspects, the disclosed technology includes: an autonomous driving super computer (ADSC) comprising: an enclosure that is coupled to a primary printed circuit board (PCB) of the ADSC; one or more power terminals that are coupled to the primary PCB and extend through the enclosure; a PCB positioned between the enclosure and the primary PCB, wherein the filter PCB includes one or more apertures to permit the one or more power terminals to extend through the filter PCB, and wherein the filter PCB is configured to mitigate electromagnetic interference emitted from the one or more power terminals.
Systems and techniques are provided for calibrating time-of-flight (TOF) sensors. An example method can include sending, from a TOF sensor system coupled to a calibration assembly, a light signal to a fiber optic cable coupled to the calibration assembly; receiving, by the calibration assembly, the light signal from the fiber optic cable; diffusing the light signal via one or more diffusers on the calibration assembly; generating, by the TOF sensor system, one or more measurements based on the diffused light signal; and based on the one or more measurements, determining one or more calibration values configured to compensate for one or more errors in the one or more measurements.
Disclosed are embodiments for facilitating automatic slice discovery and slice tuning for data mining in autonomous systems. In some aspects, an embodiment includes providing, by a processing device hosting a slice discovery machine learning (ML) model, input data to the slice discovery ML model, the input data corresponding to performance data of an autonomous vehicle (AV); identifying, by the slice discovery ML model, attributes of the AV and corresponding thresholds for the attributes that define a slice comprising a collection of data sharing common characteristics; and providing, by the slice discovery ML model, the attributes and the corresponding thresholds defining the slice to a slice miner to mine training data corresponding to the slice for a tailored dataset.
Disclosed are embodiments for facilitating generative artificial intelligence (AI) to pre-train and fine-tune models for multiple autonomous vehicle (AV) trajectories generation. In some aspects, an embodiment includes training a teacher generative AI model on a first set of training data; providing a student generative AI model with at least one distillation of the teacher generative AI model, the at least one distillation comprising transformer weights, embeddings, or predictions labels of the teacher generative AI model; training the student generative AI model that is initialized with the at least one distillation of the teacher generative AI model, wherein the student generative AI model is trained using a second set of training data that is smaller than the first set of training data; and deploying the student generative AI model to a resource-constrained environment.
Disclosed are embodiments for facilitating generative artificial intelligence to generate multiple autonomous vehicle future trajectories. In some aspects, an embodiment includes receiving input data to a generative pre-trained transformer (GPT)-based trajectory generation model, wherein the input data comprises vector map representations, nearby actor history, and autonomous vehicle (AV) history of an AV; generating map tokens from the vector map representations and generating agent tokens from the nearby actor history and the AV history; inputting a concatenated set of the map tokens and the agent tokens into an encoder transformer of the GPT-based trajectory generation model; outputting, by the encoder transformer, an output embedding that is representative of a scene of the AV; and determining, by a decoder of the GPT-based trajectory generation model, a sequence of AV waypoint predictions for the AV based on the output embedding.
Systems, methods, and non-transitory computer readable mediums are provided for providing autonomous vehicle (AV) state information of an AV to a data center when an AV state change occurs so that the data center may determine one or more interpolated AV states of the AV prior to the AV state change. For example, a system may obtain sensor data generated by one or more sensor systems of the AV. Additionally, the system may determine a first AV state change of the AV at a first time based on the sensor data. Moreover, the system may generate a first message including a portion of the sensor data associated with the kinematic change in response to determining the first AV state change. Further, the system may transmit the first message to the data center.
Systems, methods, and non-transitory computer readable mediums are provided for projecting one or more content items onto a surface outside of an AV. For example, a system may receive routing data including routing information and road agent information of each of one or more road agents within a predetermined distance threshold of an autonomous vehicle (AV). Additionally, the system may generate projection data including a content item associated with the routing information and the road agent information. Moreover, the system may provide projection data to a projection controller device communicating with one or more projection devices. The projection controller device may cause at least one of the one or more projection devices to project the content item onto a surface outside of the AV.
Systems and methods for encrypting and streaming AV sensor data and AV derived data. In particular, systems and methods are provided for receiving AV sensor data at a streaming platform and providing the received data to multiple endpoints. In some examples, the multiple endpoints are multiple viewers viewing the AV sensor data. In various examples, an AV transmits the AV sensor data to the streaming platform, and the AV sensor data can be transmitted from the streaming platform to multiple endpoints without any effect on AV central processing unit or bandwidth constraints. The AV sensor data can include multiple data streams that are layered and compressed for efficient streaming, and the streaming platform can receive real-time streamed data from the AV even as bandwidth and connection constraints vary. In various examples, the bit rate of the video ingestion from the AV can be changed based on bandwidth availability.
H04L 67/12 - Protocoles spécialement adaptés aux environnements propriétaires ou de mise en réseau pour un usage spécial, p. ex. les réseaux médicaux, les réseaux de capteurs, les réseaux dans les véhicules ou les réseaux de mesure à distance
G08G 1/00 - Systèmes de commande du trafic pour véhicules routiers
H04L 47/25 - Commande de fluxCommande de la congestion le débit étant modifié par la source lors de la détection d'un changement des conditions du réseau
Various technologies described herein pertain to systems and methods for mitigating radar signal power fading. In one embodiment, a diversity of radar sensor units having different vertical positions, or an offset, are provided. The different vertical positions introduce a phase shift to the received radar signal(s) so that when one radar sensor unit is receiving a destructive combination of signals, the other radar sensor unit is receiving a constructive combination of signals. This reduces the effect of signal power fade due to, among other things, ground bounce of radar signals destructively interfering with a radar sensor unit's ability to detect an object.
An apparatus for facilitating image segmentation on a dataset comprising a plurality of classes is described and includes a module executable by a processor to preprocess the dataset by applying a filter to labels of the plurality of classes to expand the labels of the selected class; down-sample the preprocessed dataset to a desired resolution for training an image segmentation model; and output the down-sampled preprocessed dataset to the image segmentation module, wherein the labels comprise one-hot labels.
G06V 10/26 - Segmentation de formes dans le champ d’imageDécoupage ou fusion d’éléments d’image visant à établir la région de motif, p. ex. techniques de regroupementDétection d’occlusion
G06T 3/40 - Changement d'échelle d’images complètes ou de parties d’image, p. ex. agrandissement ou rétrécissement
G06V 10/34 - Lissage ou élagage de la formeOpérations morphologiquesSquelettisation
G06V 10/764 - Dispositions pour la reconnaissance ou la compréhension d’images ou de vidéos utilisant la reconnaissance de formes ou l’apprentissage automatique utilisant la classification, p. ex. des objets vidéo
G06V 10/774 - Génération d'ensembles de motifs de formationTraitement des caractéristiques d’images ou de vidéos dans les espaces de caractéristiquesDispositions pour la reconnaissance ou la compréhension d’images ou de vidéos utilisant la reconnaissance de formes ou l’apprentissage automatique utilisant l’intégration et la réduction de données, p. ex. analyse en composantes principales [PCA] ou analyse en composantes indépendantes [ ICA] ou cartes auto-organisatrices [SOM]Séparation aveugle de source méthodes de Bootstrap, p. ex. "bagging” ou “boosting”
G06V 20/58 - Reconnaissance d’objets en mouvement ou d’obstacles, p. ex. véhicules ou piétonsReconnaissance des objets de la circulation, p. ex. signalisation routière, feux de signalisation ou routes
G06V 20/70 - Étiquetage du contenu de scène, p. ex. en tirant des représentations syntaxiques ou sémantiques
43.
UNKNOWN DRIVING HAZARD DETECTION AND RESPONSE SYSTEM
A system installed on an autonomous vehicle (AV) is described and includes an unknown driving hazard detection module configured to detect an unusual driving behavior of at least one road participant operating in proximity to the AV, wherein the unusual driving behavior comprises a deviation from a behavior of the at least one road participant as predicted by the AV; and an unknown driving hazard response module configured to cause an action to be performed in connection with the AV based on the detected unusual driving behavior.
B60W 60/00 - Systèmes d’aide à la conduite spécialement adaptés aux véhicules routiers autonomes
B60W 50/00 - Détails des systèmes d'aide à la conduite des véhicules routiers qui ne sont pas liés à la commande d'un sous-ensemble particulier
G01C 21/28 - NavigationInstruments de navigation non prévus dans les groupes spécialement adaptés pour la navigation dans un réseau routier avec corrélation de données de plusieurs instruments de navigation
44.
COMBINED WALKING AND DRIVING ROUTES FOR AUTONOMOUS VEHICLE RIDES
A walking path database stores data describing a set of connected walking paths. The walking path database may be generated based on an existing map database describing connected roads. The walking path database and map database are used to generate a combined walking and driving route, where a user walks to a particular pickup location, and a vehicle drives the user from the pickup location to a particular destination. The combined walking and driving route may be a route that minimizes the overall trip time.
A system for provisioning a plurality of resources in connection with a service to be provided in an autonomous vehicle (AV) infrastructure environment is described and includes a cloud platform executing a platform as a service (PaaS) cluster; a service-specific file specifying the plurality of resources; a values file specifying configuration information for each of the plurality of resource specified in the service-specific file; and a service for deploying the plurality of resources comprising the service on the cloud services platform using the service-specific file and the values file.
B60W 60/00 - Systèmes d’aide à la conduite spécialement adaptés aux véhicules routiers autonomes
H04L 67/12 - Protocoles spécialement adaptés aux environnements propriétaires ou de mise en réseau pour un usage spécial, p. ex. les réseaux médicaux, les réseaux de capteurs, les réseaux dans les véhicules ou les réseaux de mesure à distance
46.
Voice pre-processing pipeline for exterior communications on autonomous vehicle
Systems and methods for vehicles to communicate with people outside the vehicle. In particular, a two-way communication system is provided including exterior microphone arrays on the vehicle. Calls performed from the exterior of a vehicle can be difficult to implement with acceptable quality in noisy environments. A pipeline having multiple beamformers is provided to improve call quality, where the pipeline operates on the incoming data for each of the microphone arrays. Each beamformer estimates direction of arrival (DOA) of the speech signal, and utilizes the estimated DOA to generate spatial filtering coefficients to filter the signal. A beamformer can be used to generate a one channel output for each microphone array, and a mixer can be used to mix the signal output from each microphone array. The beamformer enables reliable two-way communication with acceptable listening quality and good intelligibility for the remote assistant listening to the captured signal.
H04R 1/40 - Dispositions pour obtenir la fréquence désirée ou les caractéristiques directionnelles pour obtenir la caractéristique directionnelle désirée uniquement en combinant plusieurs transducteurs identiques
Systems and techniques are provided for an autonomous vehicle (AV) to select a mobile network. An example method can include establishing a connection between the AV and a mobile device associated with a passenger of the AV; identifying, using the connection, a first mobile network associated with the mobile device; determining that the first mobile network associated with the mobile device is different than a second mobile network associated with the AV; determining, based on a first network indicator associated with the first mobile network and a second network indicator associated with the second mobile network, that the first mobile network is preferable to the second mobile network; and initiating a mobile connection between the AV and the first mobile network.
Systems and techniques are provided for determining simulation fidelity. An example method includes receiving, by a machine learning model, a plurality of input datasets that are generated as part of a simulated scene within a simulation environment; generating, by the machine learning model, a plurality of outputs that are based on the plurality of input datasets; and determining, by a discriminator head of the machine learning model, a plurality of input classifiers for each of the plurality of outputs, wherein each input classifier from the plurality of input classifiers indicates whether input data corresponding to a respective output from the plurality of outputs is associated with simulated input data or real-world input data.
The present disclosure generally relates to improved autonomous vehicle (AV) navigation in foggy conditions and, more specifically, to determining a fog intensity level and adjusting the speed of the AV based on the fog intensity level. In some aspects, a method of the disclosed technology includes steps for collecting sensor data for an environment around an AV; determining, based on the collected sensor data, that fog exists in the environment around the AV; determining, based on the collected sensor data, a fog proxy level; determining, based on the fog proxy level, a track spawning range of the AV; and adjusting, based on the track spawning range of the AV, a speed of the AV. Systems and machine-readable media are also provided.
Systems and techniques are provided for dynamically implementing vehicle sensor configurations based on operational contexts. An example method can include detecting, based on at least one of map data and sensor data from at least one sensor on a vehicle, one or more characteristics of an operational context of the vehicle; determining, based on the one or more characteristics of the operational context, a sensor configuration for the vehicle to implement when navigating the operational context, the sensor configuration comprising one or more sensors selected for use by the vehicle in the operational context and one or more different sensors set to an off state or a reduced operating mode while the vehicle is in the operational context; and dynamically implementing the sensor configuration at the vehicle when the vehicle is in the operational context.
Systems and techniques are provided for synchronizing sensor operations. An example method includes determining a scanning frequency of a light detection and ranging (LIDAR) sensor configured to collect data for regions of space during each scan cycle; selecting an exposure from an exposure sequence generated based on data captured by a time-of-flight (TOF) sensor to align with data from a scan from the LIDAR sensor during a scan cycle; based on the scanning frequency, a field-of-view (FOV) of the LIDAR sensor, a FOV of the TOF sensor, a location of the exposure within the exposure sequence, and/or sensor internal delays, determining a timeframe between a reference time and an alignment time during the scan cycle when the FOVs of the LIDAR sensor and the TOF sensor are aligned; and based on the timeframe, determining a time offset for triggering the TOF sensor to capture data associated with the exposure sequence.
Systems and techniques are provided for visualizing a driving scene overlaid with predictions. An example process includes receiving sensor data collected by one or more sensors of an autonomous vehicle (AV) navigating in a scene, receiving at least one prediction associated with one or more objects in the scene, and generating a superimposed visualization representing the AV in the scene and representing the at least one prediction associated with the one or more objects in the scene. The superimposed visualization may dynamically display the at least one prediction associated with the one or more objects in time sequence as the AV navigates in the scene.
Disclosed are embodiments for facilitating right-of-way-based semantic coverage and automatic labeling for trajectory generation in autonomous systems. In some aspects, an embodiment includes receiving a set of trajectories generated by a trajectory generation source, the set of trajectories generated for an autonomous vehicle (AV) interaction with a road agent; labeling a ground truth trajectory from the set of trajectories with a ground truth label; classifying each remaining trajectory of the set of trajectories as at least one of an assert trajectory or a yield trajectory; for an assert group comprising the assert trajectories, assigning an auxiliary assert label to the assert trajectory having a highest selection score; for a yield group comprising the yield trajectories, assigning an auxiliary yield label to the yield trajectory having a highest selection score; and utilizing the ground truth label, the auxiliary assert label, and the auxiliary yield label to train a trajectory generation model.
Disclosed are embodiments for facilitating autonomous vehicle (AV) pullover clustering prevention. In some aspects, an embodiment includes receiving identification of an origin location and a destination location corresponding to a transportation trip request for an AV; determining a set of pullover locations comprising pickup locations for the origin location and drop-off locations for the destination location; for each pullover location of the set of pullover locations: determining an estimated time of arrival (ETA) time window for the AV at the pullover location; determining a number of other AVs expected to be at the pullover location during the ETA time window; and responsive to the number of other AVs expected to be at the pullover location during the ETA time window exceeding an AV pullover crowding metric, removing the pullover location from the set of pullover locations to produce a revised set of pullover locations.
G08G 1/00 - Systèmes de commande du trafic pour véhicules routiers
B60W 40/00 - Calcul ou estimation des paramètres de fonctionnement pour les systèmes d'aide à la conduite de véhicules routiers qui ne sont pas liés à la commande d'un sous-ensemble particulier
B60W 60/00 - Systèmes d’aide à la conduite spécialement adaptés aux véhicules routiers autonomes
G06Q 50/47 - Requêtes de disponibilité de places de passager pour effectuer un trajet, p. ex. sollicitation d’une course
A system includes an electrical component, a sensor system, and a connector. The sensor system has a housing with a connector opening and circuitry disposed within the housing. The connector is disposed within the connector opening of the housing of the sensor system and electrically connects the electrical component to the circuitry of the sensor system. The connector includes a vent that allows for movement of air into and out of the housing via the connector opening while preventing moisture from moving into the housing of the sensor system.
Disclosed herein is a radar target simulator (RTS) device including a processor and a memory. The memory stores computer-readable instructions that, when executed by the processor, cause the processor to perform acts including receiving a test sensor probe signal from a test sensor. The acts further include generating a first virtual target for a radar target simulation and transmitting a virtually reflected waveform responsive to the test sensor probe signal for the radar target test simulation, wherein the virtually reflected waveform is based at least in part on a change in a position of the first virtual target.
Disclosed are embodiments for facilitating data mining on an edge platform using repurposed neural network models in autonomous systems. In some aspects, an embodiment includes receiving, by a processing device hosting a data source proxy head of a machine learning (ML) model deployed on an autonomous vehicle (AV), a set of features selected from raw data by a backbone network of the ML model; utilizing, by the data source proxy head, the set of features selected from the raw data as input data to a trained data source mining model of the data source proxy head; identifying, by the trained data source mining model based on the input data, a portion of the raw data to classify as mining data; and providing, by the data source proxy head, identification of the portion of the raw data as a data mining output.
G05B 13/02 - Systèmes de commande adaptatifs, c.-à-d. systèmes se réglant eux-mêmes automatiquement pour obtenir un rendement optimal suivant un critère prédéterminé électriques
B60W 50/00 - Détails des systèmes d'aide à la conduite des véhicules routiers qui ne sont pas liés à la commande d'un sous-ensemble particulier
B60W 60/00 - Systèmes d’aide à la conduite spécialement adaptés aux véhicules routiers autonomes
Systems and methods for providing opportunities for interactions and communication between vehicles and passengers. In particular, a vehicle socialization platform is provided for facilitating a wide range of social and utilitarian exchanges between vehicles and their occupants, thus providing a community of interconnected vehicles. Vehicle interactions can include passive communication, including vehicles from a particular fleet acknowledging other fleet vehicles without intervention from a passenger or user. Vehicle interactions can include active communication, in which passengers are engaged with and have some control over vehicle communications through dedicated experiences provided in a ridehail application or on an in-vehicle tablet. In particular, some experiences can become dynamically available based on the fleet vehicle's proximity to another fleet vehicle. The vehicle socialization platform is a smart socialization platform that enables two-way communication between vehicles and passengers, facilitating a wide range of social and utilitarian exchanges between vehicles and vehicle occupants.
G05D 1/246 - Dispositions pour déterminer la position ou l’orientation utilisant des cartes d’environnement, p. ex. localisation et cartographie simultanées [SLAM]
G05D 1/646 - Suivi d’une trajectoire prédéfinie, p. ex. d’une ligne marquée sur le sol ou d’une trajectoire de vol
A radar apparatus at a vehicle may transmit a set of radar energy that is reflected off an object after which the reflected radar energy may be received at the radar apparatus. When relative motion of a radar apparatus and an object include movement in two different directions, power of the reflected radar signals may be spread out in a manner that makes data associated with the reflected radar signals difficult to interpret. This spreading out of the radar signals can make mappings of the received radar data appear to be smeared or distorted. To compensate for this smearing effect, mappings of this smeared radar data may be compared with curves from which compensation factors may be identified. These compensation factors may allow a processor to perform calculations to generate updated mappings of the radar data that may allow the processor to more accurately identify characteristics of the object.
Systems, methods, and non-transitory computer readable mediums are provided for detecting traffic lights in images based on prompts that automatically limit the search space within the images for the traffic lights. For example, a system may receive image data and prompt data identifying a search space for traffic light detection. Additionally, the system may generate layout data based on the prompt data and generate traffic light data based on the image data and the layout data. Moreover, the system may determine state information for one or more first traffics included in the image data based on the traffic light data. Further, the system may implement one or more operations associated with a navigation and routing system of the autonomous vehicle based on the state information.
B60W 60/00 - Systèmes d’aide à la conduite spécialement adaptés aux véhicules routiers autonomes
G01C 21/36 - Dispositions d'entrée/sortie pour des calculateurs embarqués
G06T 7/73 - Détermination de la position ou de l'orientation des objets ou des caméras utilisant des procédés basés sur les caractéristiques
G06V 20/58 - Reconnaissance d’objets en mouvement ou d’obstacles, p. ex. véhicules ou piétonsReconnaissance des objets de la circulation, p. ex. signalisation routière, feux de signalisation ou routes
A radar sensor system comprises a receive antenna that receives a radar return and circuitry configured to perform certain acts. The acts comprise receiving analog-to-digital converted (ADC) samples of a received radar signal and identifying power magnitudes for the ADC samples. The acts also comprise calculating a mean power value for ADC samples in a given ramp of the received radar signal and calculating a mean power variance of the ADC samples in the given ramp. The acts further comprise identifying a sample having a power magnitude above a power threshold. Additionally, the acts comprise adjusting the power magnitude of the identified sample to mitigate the interference effect caused by the identified sample.
Various technologies described herein pertain to a radar sensor system that determines if a point detected by the radar corresponds to a true target and outputs an indication upon which control of an autonomous vehicle may be based. The radar can be used to determine if the point corresponds to a false target, thereby mitigating false positives. The radar can be used to determine if an object coincides with a location of the false target, thereby mitigating false negatives. The radar employs an algorithm that includes applying different windows to a return signal on a beamforming stage and comparing the resulting windowed signals.
G01S 7/41 - Détails des systèmes correspondant aux groupes , , de systèmes selon le groupe utilisant l'analyse du signal d'écho pour la caractérisation de la cibleSignature de cibleSurface équivalente de cible
G01S 7/02 - Détails des systèmes correspondant aux groupes , , de systèmes selon le groupe
G01S 13/931 - Radar ou systèmes analogues, spécialement adaptés pour des applications spécifiques pour prévenir les collisions de véhicules terrestres
63.
Restarting operations for recovery processes in autonomous systems
Disclosed are embodiments for facilitating restarting operations for recovery processes in autonomous systems. In some aspects, an embodiment includes determining that an autonomous vehicle (AV) is experiencing a failure condition; identifying a plurality of restart operations to apply to one or more components of the AV contributing to the failure condition; prior to applying a restart operation of the plurality of restart operations, determining that safety conditions corresponding to the restart operation are satisfied; and applying the plurality of restart operations to the AV in accordance with an increasing order of disruptiveness of each of the restart operations to operations of the AV.
B60W 50/02 - Détails des systèmes d'aide à la conduite des véhicules routiers qui ne sont pas liés à la commande d'un sous-ensemble particulier pour préserver la sécurité en cas de défaillance du système d'aide à la conduite, p. ex. en diagnostiquant ou en palliant à un dysfonctionnement
B60W 60/00 - Systèmes d’aide à la conduite spécialement adaptés aux véhicules routiers autonomes
64.
Display screen or portion thereof with a graphical user interface
Systems, methods, and apparatuses are provided for monitoring and automatically limiting an electrical current flowing to a pulsed or switching component that is voltage driven or controlled. For example, an apparatus may include a transistor electrically coupled to a power supply. In some examples, the transistor being configured to control current flow to a driver component. Additionally, the apparatus may include a resistor electrically coupled to the driver component, and a current monitor electrically coupled to a gate of the transistor and the resistor. In some examples, the current monitor may be configured to receive a threshold voltage, and receive an output voltage. Additionally, the current monitor may be configured to determine, whether the threshold voltage is greater than the output voltage, and based on the determination of whether the threshold voltage is greater than the output voltage, adjust the transistor to control the current flow to the driver component.
H03K 17/56 - Commutation ou ouverture de porte électronique, c.-à-d. par d'autres moyens que la fermeture et l'ouverture de contacts caractérisée par l'utilisation de composants spécifiés par l'utilisation, comme éléments actifs, de dispositifs à semi-conducteurs
B60R 16/03 - Circuits électriques ou circuits de fluides spécialement adaptés aux véhicules et non prévus ailleursAgencement des éléments des circuits électriques ou des circuits de fluides spécialement adapté aux véhicules et non prévu ailleurs électriques pour l'alimentation des sous-systèmes du véhicule en énergie électrique
G01R 19/165 - Indication de ce qu'un courant ou une tension est, soit supérieur ou inférieur à une valeur prédéterminée, soit à l'intérieur ou à l'extérieur d'une plage de valeurs prédéterminée
Disclosed are embodiments for facilitating real-time autonomous vehicle (AV) fleet parking availability. In some aspects, an embodiment includes determining, based on a first set of perception sensor inputs of an AV, sections of roadway that are used for vehicle parking; determining, based on a second set of perception sensor inputs, whether the sections of roadway are allowable for parking and are available for parking; transmitting parking location data comprising identification of parking location objects corresponding to the sections of roadway and signals indicating that the parking location object is used for vehicle parking, indicating whether the parking location object is allowable for parking, and indicating whether the parking location object is available for parking, wherein the parking location data is aggregated with other parking location data from other AVs into aggregated parking location data; and identifying a location for parking of the AV based on the aggregated parking location data.
Sensor data obtained from vehicles driving through a particular environment (e.g., a particular city or region being mapped) is used to identify and label traffic control features. Location, speed, and/or acceleration of mapping vehicles can be used to identify intersections that may have traffic control features. Environmental data, such as image data captured by one or more cameras, and point cloud data collected by lidar and/or radar sensors, is used to automatically detect traffic control features at the intersections.
G06V 20/70 - Étiquetage du contenu de scène, p. ex. en tirant des représentations syntaxiques ou sémantiques
B60W 60/00 - Systèmes d’aide à la conduite spécialement adaptés aux véhicules routiers autonomes
G01C 21/00 - NavigationInstruments de navigation non prévus dans les groupes
G06T 7/73 - Détermination de la position ou de l'orientation des objets ou des caméras utilisant des procédés basés sur les caractéristiques
G06V 10/80 - Fusion, c.-à-d. combinaison des données de diverses sources au niveau du capteur, du prétraitement, de l’extraction des caractéristiques ou de la classification
G06V 20/56 - Contexte ou environnement de l’image à l’extérieur d’un véhicule à partir de capteurs embarqués
G06V 20/58 - Reconnaissance d’objets en mouvement ou d’obstacles, p. ex. véhicules ou piétonsReconnaissance des objets de la circulation, p. ex. signalisation routière, feux de signalisation ou routes
G08G 1/01 - Détection du mouvement du trafic pour le comptage ou la commande
Planning, or path finding is one of many tasks performed by an autonomous vehicle (AV). Planning involves searching for and updating an optimal plan from a current pose to an end pose while avoiding obstacles. Planning can be particularly challenging in driving scenarios involving complex driving maneuvers, such as parking, making a U-turn, making a K-turn, etc. Searching for an optimal plan may take more time, e.g., a few seconds, in certain driving scenarios. Consumers of the plan may have a low tolerance for delay in receiving the optimal plan. A relay can be implemented to publish the latest plan to the consumers at a frequency. The relay can check whether the delay violates a dynamic threshold and raise an error accordingly. The searching process can be improved to reduce the delay in finding an optimal plan.
Disclosed are embodiments for facilitating a multi-inertial measurement unit (IMU) combination unit. In some aspects, an embodiment includes receiving, at a multi-IMU combination unit (MICU), sensor data from a plurality of inertial data sensors of a same sensor type; for each inertial data sensor, calibrating and transforming the respective sensor data using a calibration estimate for the inertial data sensor, where the calibration estimate is based on pre-integration methods that provide individual kinematic feedback that is compared to fused kinematic feedback from a main filter; combining the calibrated and transformed sensor data from the plurality of inertial data sensors into a fused output for the same sensor type; sampling the fused output to provide a single inertial data measurement for the plurality of inertial data sensors; and providing the fused kinematic feedback for the calibration estimate, the fused kinematic feedback generated from the sampling of the single fused output.
G01C 25/00 - Fabrication, étalonnage, nettoyage ou réparation des instruments ou des dispositifs mentionnés dans les autres groupes de la présente sous-classe
70.
DETECTING SENSOR DEFECTS FOR THREE-DIMENSIONAL TIME-OF-FLIGHT SENSORS
Described herein are systems and techniques for evaluating whether components of a sensing device are functioning properly. An example method includes obtaining a first set of measurements corresponding to a pixel from a first image frame captured by a sensor, wherein each respective measurement from the first set of measurements is associated with a different signal phase; converting the first set of measurements using an analog to digital converter (ADC) to yield a first set of digitized measurements; and determining, based on the first set of digitized measurements, a first confidence map value that corresponds to the pixel from the first image frame, wherein the first confidence map value is indicative of at least one defect in the sensor.
Systems, methods, and apparatuses are provided for testing and measuring the behavior or logic of one or more cloud services associated with an updated AV bot script. For example, a system may be configured to receive a first autonomous vehicle (AV) bot script, reference data associated with a first cloud service, and configuration data identifying at least a first variable of a set of adjustable variables of a simulated AV, a first value of a set of values associated with the first variable, and the first cloud service. Additionally, the system may be configured to generate a second AV bot script based on the configuration data and the first AV bot script. Further, the system may be configured to receive cloud output data associated with the first cloud service and determine one or more differences between the cloud output data and the reference data.
A radar system comprises a radar sensor and circuitry configured to perform acts comprising receiving radar sensor data from the radar sensor and detecting static objects in the radar sensor data. The acts further comprise calculating cell parameters related to the static objects for a given radar cell, the cell parameters including a minimum residual velocity value for a static object detection. The acts also comprise generating a 2D cell map grid representing 3D radar cells including the given radar cell and comparing the calculated cell parameters to respective threshold values. Additionally, the acts comprise assigning weight scores to the cell parameters based on the threshold comparison and summing the weight scores assigned to the cell parameters to generate a combined score. The acts also comprise comparing the combined score to a probability threshold and outputting an indication that the cell is occupied or not occupied.
G01S 7/41 - Détails des systèmes correspondant aux groupes , , de systèmes selon le groupe utilisant l'analyse du signal d'écho pour la caractérisation de la cibleSignature de cibleSurface équivalente de cible
G01S 13/536 - Discrimination entre objets fixes et mobiles ou entre objets se déplaçant à différentes vitesses utilisant la transmission d'ondes continues non modulées, ou modulées en amplitude, en fréquence ou en phase
G01S 13/89 - Radar ou systèmes analogues, spécialement adaptés pour des applications spécifiques pour la cartographie ou la représentation
Systems and methods for a centrally designed vehicle HVAC system that cools vehicle components as well as the cabin and/or delivery containers. The HVAC system is designed with the ability to divert HVAC capacity from one vehicle system to another, allowing for more optimized HVAC usage. In some examples, the HVAC system disclosed herein allows for a more optimized HVAC system size, including decreased HVAC system weight and power consumption. In various implementations, the HVAC system uses back office prediction and remote control to make the determination of HVAC system bias based on current weather conditions as well as forecasted weather conditions. The HVAC system provides a comfortable vehicle cabin while ensuring the thermal condition of vehicle systems and components are managed to avoid interruptions to service.
B60H 1/00 - Dispositifs de chauffage, de refroidissement ou de ventilation
B60H 1/14 - Dispositifs de chauffage, de refroidissement ou de ventilation la chaleur étant prélevée de l'installation de propulsion autrement que par le liquide de refroidissement de l'installation
74.
Identification and fusion of super pixels and super voxels captured by time-of-flight sensors
Systems and techniques are provided for processing image data from a time-of-flight sensor. An example method includes receiving a depth map and a two-dimensional image that each correspond to an image frame captured by a time-of-flight (ToF) sensor; identifying a plurality of pixels within the two-dimensional image that correspond to at least one scene element; identifying, based on the plurality of pixels, a plurality of voxels within the depth map that correspond to the at least one scene element; and grouping the plurality of voxels into a super voxel that corresponds to the at least one scene element.
G06T 7/521 - Récupération de la profondeur ou de la forme à partir de la télémétrie laser, p. ex. par interférométrieRécupération de la profondeur ou de la forme à partir de la projection de lumière structurée
G06V 10/70 - Dispositions pour la reconnaissance ou la compréhension d’images ou de vidéos utilisant la reconnaissance de formes ou l’apprentissage automatique
G06V 10/80 - Fusion, c.-à-d. combinaison des données de diverses sources au niveau du capteur, du prétraitement, de l’extraction des caractéristiques ou de la classification
G06V 20/58 - Reconnaissance d’objets en mouvement ou d’obstacles, p. ex. véhicules ou piétonsReconnaissance des objets de la circulation, p. ex. signalisation routière, feux de signalisation ou routes
75.
STATISTICALLY MODELING EFFECT OF FOG ON LIDAR DATA
The systems and methods disclosed herein address simulating the effect of fog on a photon. One method defines a target at a position in a 3D environment and includes the steps of selecting a starting position of the photon in the 3D environment, selecting a propagation vector directed from the starting position toward the target, selecting a propagation distance, determining a new position of the photon based in part on the starting position of the photon and the propagation vector and the propagation distance, determining whether the photon is absorbed before reaching the new position and determining, if the photon has not been absorbed, whether the photon intersects the target before reaching the new position.
Systems and methods of simulating an effect of fog on a Light Detection and Ranging (LiDAR) sensor are disclosed. The method includes the steps of determining whether a target is present within the field-of-view (FOV) of the LiDAR sensor, determining a fog probability density function (PDFfog) vs range, modifying, if a target is present within the FOV, the PDFfog to account for the target, calculating a cumulative density function (CDF) for the PDFfog, randomly sampling the CDF to determine a plurality of ranges and additively plotting a predetermined Gaussian distribution centered on each range, and identifying a peak value of the additive plot and reporting the range associated with the peak value as the strongest return of the LiDAR unit.
System, methods, and computer-readable media for an object path prediction model, and an associated training technique, to output a path that is considered an object collision path. The object path prediction model outputs a set of predicted paths for the object that are outputted to a trained planning algorithm, which includes paths that are most likely to occur and a path that is considered an object collision path. The predicted paths are sent to the trained planning algorithm and used for planning a trajectory for the autonomous vehicle that is associated with a low probability of colliding with or taking a sudden evasive action to avoid the object.
B60W 30/095 - Prévision du trajet ou de la probabilité de collision
B60W 40/04 - Calcul ou estimation des paramètres de fonctionnement pour les systèmes d'aide à la conduite de véhicules routiers qui ne sont pas liés à la commande d'un sous-ensemble particulier liés aux conditions ambiantes liés aux conditions de trafic
B60W 60/00 - Systèmes d’aide à la conduite spécialement adaptés aux véhicules routiers autonomes
G06V 10/70 - Dispositions pour la reconnaissance ou la compréhension d’images ou de vidéos utilisant la reconnaissance de formes ou l’apprentissage automatique
G06V 20/58 - Reconnaissance d’objets en mouvement ou d’obstacles, p. ex. véhicules ou piétonsReconnaissance des objets de la circulation, p. ex. signalisation routière, feux de signalisation ou routes
The present disclosure generally relates to autonomous vehicles, and, more specifically, to systems and techniques for extending a sensor's capabilities by extending the field of view into one or more blind spots located in the environment of an autonomous vehicle. In some aspects, a method of the disclosed technology includes: determining a location of a first blind spot that is within range of a LIDAR sensor; positioning the reflective object according to a first pose; triggering the LIDAR sensor to collect sensor data during a scan cycle performed by the LIDAR sensor, wherein at least one portion of a light beam emitted from the LIDAR sensor is reflected off the reflective object at the first pose toward the location of the first blind spot. Systems and machine-readable media are also provided.
An AV may detect degraded states of other AVs. For instance, the AV may use its sensors to detect one or more other vehicles in the local area. The AV may determine whether any of these vehicles is a peer of the AV based on features of these vehicles captured by its sensors. Alternatively, the AV may determine if any vehicle is a peer based on encrypted communication with the vehicle. The AV may also determine whether any peer AV has a state that deviates from an expected or desirable state. The AV may communicate with an online system and request the online system to provide information for identifying the vehicle or detecting degraded states. The AV may also report degradations in peer AVs to the online system. The online system may recover the degraded AV or dispatch another AV to complete a task assigned to the degraded AV.
G07C 5/00 - Enregistrement ou indication du fonctionnement de véhicules
G07C 5/08 - Enregistrement ou indication de données de marche autres que le temps de circulation, de fonctionnement, d'arrêt ou d'attente, avec ou sans enregistrement des temps de circulation, de fonctionnement, d'arrêt ou d'attente
H04W 4/46 - Services spécialement adaptés à des environnements, à des situations ou à des fins spécifiques pour les véhicules, p. ex. communication véhicule-piétons pour la communication de véhicule à véhicule
The present disclosure generally relates to improving the accuracy of autonomous vehicle simulations by identifying which test results are most dependent on a hardware type and/or software version used in the computing systems of the simulation and the AV. In some aspects, a method of the disclosed technology includes steps for performing a test on a first hardware component type associated with an AV to produce a first output; performing the test on a second hardware component type associated with a simulation of an AV to produce a second output; comparing the first output with the second output using a statistical analysis to determine a value related to a difference between the first output and the second output; and assigning a weight to the test based on the value related to the difference between the first output and the second output. Systems and machine-readable media are also provided.
G06F 30/15 - Conception de véhicules, d’aéronefs ou d’embarcations
G06F 30/27 - Optimisation, vérification ou simulation de l’objet conçu utilisant l’apprentissage automatique, p. ex. l’intelligence artificielle, les réseaux neuronaux, les machines à support de vecteur [MSV] ou l’apprentissage d’un modèle
81.
PREDICTION OF MOVABILITY OF AN UNCLASSIFIED OBJECT
Systems and techniques are provided for predicting a movability of an unclassified object. An example process includes receiving sensor data captured within a single frame, identifying an unclassified object in the sensor data, and providing the sensor data to a neural network, which is configured to predict a motion signal for the unclassified object in the scene. The example process can further include determining whether the unclassified object is a static object or a dynamic object based on the motion signal.
B60W 60/00 - Systèmes d’aide à la conduite spécialement adaptés aux véhicules routiers autonomes
B60W 50/00 - Détails des systèmes d'aide à la conduite des véhicules routiers qui ne sont pas liés à la commande d'un sous-ensemble particulier
G06V 10/26 - Segmentation de formes dans le champ d’imageDécoupage ou fusion d’éléments d’image visant à établir la région de motif, p. ex. techniques de regroupementDétection d’occlusion
G06V 10/82 - Dispositions pour la reconnaissance ou la compréhension d’images ou de vidéos utilisant la reconnaissance de formes ou l’apprentissage automatique utilisant les réseaux neuronaux
G06V 20/56 - Contexte ou environnement de l’image à l’extérieur d’un véhicule à partir de capteurs embarqués
A unified electronic control unit (ECU) interface provides a bridge from ECU clients to the ECU. The ECU interface may be implemented on a vehicle's onboard computer, and it may be used by clients executing on the onboard computer and, in some cases, by clients outside the onboard computer. The ECU interface may receive a request from a client in a first messaging format, e.g., a local networking protocol used by processes running on the vehicle. The ECU interface translates the request into a second messaging format and transmits the request to the ECU. The ECU interface receives a response from the ECU and translates the response into the first messaging format for delivery to the client. The ECU interface may further perform authentication of the client and/or the ECU, packet serialization of the response from the ECU, packet diagnostics, and other functionalities associated with communicating with the ECU.
Systems and techniques are provided for providing spatial path guidance of an autonomous vehicle (AV) in a remote assistance session by leveraging input from a remote operator. An example process includes initiating a remote assistance session between a remote assistant and an AV in a scene that comprises at least a portion of a planned path of the AV, receiving a waypoint input from the remote assistant associated with the remote assistance session, and generating a spatial path for the AV to navigate based on the one or more waypoints that are provided in the waypoint input. The example process includes generating an endpoint along the spatial path and terminating the remote assistance mode in response to determining that the AV has navigated past the last control point or gracefully stopping at the endpoint to receive further assistance.
Systems and techniques are provided for a decentralized private service network for autonomous vehicles (AVs). An example method can include receiving a request to register a user premises as a private service station that AVs can use to receive maintenance, the request specifying an availability of the user premises for use to receive maintenance; in response to the request, adding the user premises to a network of service stations including a plurality of user premises registered as private service stations for use by the AVs to receive maintenance and AV service facilities operated by a commercial business/businesses; determining a distance between an AV and the user premises; and determining whether to schedule a battery of the AV to be charged at the user premises based on a charge level of the battery, the specified availability of the user premises, and the distance between the AV and the user premises.
The present disclosure generally relates to detecting and filtering self-hit data from a sensor mounted to an autonomous vehicle and, more specifically, to identifying self-hit sensor data and generating and applying an image mask to filter out the self-hit sensor data. In some aspects, a method of the disclosed technology includes steps for collecting first sensor data for an environment around an autonomous vehicle (AV); identifying one or more data points, from the collected first sensor data, that correspond with a surface of the AV; generating an image mask representing the one or more data points that correspond with the surface of the AV; collecting second sensor data for the environment around the AV; and applying the image mask to the collected second sensor data. Systems and machine-readable media are also provided.
Systems and techniques are provided for detecting a translucent matter based on light detection and ranging (LIDAR) returns. An example method can include receiving at least two LIDAR returns associated with a LIDAR beam transmitted by a LIDAR device. The two LIDAR returns include a primary return comprising a first portion of the LIDAR beam reflected from a matter and a secondary return comprising a second portion of the LIDAR beam reflected from additional matter. The example method can further include determining a distance difference between a first position of the matter along the first path and a second position of the additional matter along the second path, comparing the distance difference with a threshold, and based on the comparison between the distance difference and the threshold, determining whether the additional matter is a non-translucent matter or a translucent matter.
Systems and techniques are provided for fusing raw sensor data captured by a camera sensor and one or more depth sensors to generate a dense depth map. An example process includes receiving raw camera data captured by a camera sensor and descriptive of a scene, receiving raw depth data captured by one or more depth-sensing sensors and descriptive of the scene, and providing the raw camera data and the raw depth data to a neural network, which is configured to fuse the raw camera data and the raw depth data. The example process can further include generating a depth map of the scene based on the fusion of the raw camera data and the raw depth data.
G06V 20/58 - Reconnaissance d’objets en mouvement ou d’obstacles, p. ex. véhicules ou piétonsReconnaissance des objets de la circulation, p. ex. signalisation routière, feux de signalisation ou routes
G06T 5/60 - Amélioration ou restauration d'image utilisant l’apprentissage automatique, p. ex. les réseaux neuronaux
G06T 7/55 - Récupération de la profondeur ou de la forme à partir de plusieurs images
G06V 10/80 - Fusion, c.-à-d. combinaison des données de diverses sources au niveau du capteur, du prétraitement, de l’extraction des caractéristiques ou de la classification
G06V 10/82 - Dispositions pour la reconnaissance ou la compréhension d’images ou de vidéos utilisant la reconnaissance de formes ou l’apprentissage automatique utilisant les réseaux neuronaux
Aspects of the disclosed technology provide solutions for searching point cloud data, such as Light Detection and Ranging (LiDAR) data and in particular, for using multi-modal embeddings for searching objects within a LiDAR data set. A process of the disclosed technology can include steps for receiving road data, wherein the road data represents a real-world environment encountered by an autonomous vehicle (AV) and wherein the road data comprises point cloud data representing a plurality of objects and generating, for each of the plurality of objects, a corresponding set of first embeddings. The process can further include steps for receiving a text string corresponding to a searched object, generating a second embedding corresponding to the searched object and identifying a matching object among the plurality of objects based on a comparison of the set of first embeddings and the second embedding. System and machine-readable media are also provided.
G06F 16/583 - Recherche caractérisée par l’utilisation de métadonnées, p. ex. de métadonnées ne provenant pas du contenu ou de métadonnées générées manuellement utilisant des métadonnées provenant automatiquement du contenu
G06T 7/50 - Récupération de la profondeur ou de la forme
G06V 20/56 - Contexte ou environnement de l’image à l’extérieur d’un véhicule à partir de capteurs embarqués
G06V 20/70 - Étiquetage du contenu de scène, p. ex. en tirant des représentations syntaxiques ou sémantiques
A radar system comprises a transmit antenna that transmits a radar signal and a receive antenna that receives a radar return. The system further comprises one or more processors configured to perform acts comprising: detecting object based on information in the radar return. The acts further comprise generating and providing to a tracking system a point cloud comprising information related to the detected object. Additionally, the acts comprise receiving from the tracking system an indication of a predicted cell in which the detected object is predicted to be located in a subsequent radar frame, wherein the predicted cell is associated with a confidence value. The acts further comprise reducing, for the subsequent radar frame, a detection threshold value for the predicted cell.
G01S 7/292 - Récepteurs avec extraction de signaux d'échos recherchés
G01S 7/41 - Détails des systèmes correspondant aux groupes , , de systèmes selon le groupe utilisant l'analyse du signal d'écho pour la caractérisation de la cibleSignature de cibleSurface équivalente de cible
G01S 13/42 - Mesure simultanée de la distance et d'autres coordonnées
G01S 13/931 - Radar ou systèmes analogues, spécialement adaptés pour des applications spécifiques pour prévenir les collisions de véhicules terrestres
90.
Switchable terminations based on device identifier on can bus
A set of devices are coupled to a control area network (CAN) bus. At least some of the devices include switchable resistors that can apply a termination resistor to the CAN bus. Identification pins are used to identify devices on the CAN bus, and also to identify which device or devices are at the ends of the CAN bus. If a device is at an end of the CAN bus, the switch of the switchable resistor is set to couple the termination resistor to the CAN bus. Otherwise, the termination resistor is not coupled to the CAN bus.
B60Q 1/50 - Agencement des dispositifs de signalisation optique ou d'éclairage, leur montage, leur support ou les circuits à cet effet les dispositifs ayant principalement pour objet d'indiquer le contour du véhicule ou de certaines de ses parties, ou pour engendrer des signaux au bénéfice d'autres véhicules pour indiquer d'autres intentions ou conditions, p. ex. demandes d'attente ou de dépassement
B60R 16/023 - Circuits électriques ou circuits de fluides spécialement adaptés aux véhicules et non prévus ailleursAgencement des éléments des circuits électriques ou des circuits de fluides spécialement adapté aux véhicules et non prévu ailleurs électriques pour la transmission de signaux entre des parties ou des sous-systèmes du véhicule
91.
CHAINING MACHINE LEARNING MODELS WITH CONFIDENCE LEVEL OF AN OUTPUT
Systems and techniques are provided for chaining machine learning (ML) models using a confidence level of an output of a provider model. An example method can include receiving a first output generated by a first ML model, determining a confidence level of the first output generated by the first ML model, and providing the first output of the first ML model and the confidence level of the first output to a second ML model as an input of the second ML model. The second ML model can be configured to process the first output of the first ML model and the confidence level of the first output of the first ML model to generate a second output of the second ML model.
Aspects of the disclosed technology provide solutions for searching voxel data, such as voxelized Light Detection and Ranging (LiDAR) point cloud data and in particular, for using multi-modal embeddings for searching objects within a voxel data set. A process of the disclosed technology can include steps for receiving sensor data, wherein the sensor data represents a real-world environment encountered by an autonomous vehicle (AV) and wherein the sensor data comprises point cloud data representing a plurality of objects; generating a voxel representation for the plurality of objects; generating, based on the voxel representation for the plurality of objects, a corresponding set of first embeddings; receiving a text string; generating a second embedding corresponding to the text string; and identifying a matching object among the plurality of objects based on a comparison of the set of first embeddings and the second embedding. System and machine-readable media are also provided.
The technologies described herein relate to computing a point cloud based upon data output by several radar sensors in a distributed radar system. The radar sensors generate tensors based upon echo signals detected by the radar sensors. Values are extracted from the tensors and a sequence of tokens is created, where the sequence of tokens includes the values extracted from the tensors. The sequence of tokens is provided as input to a transformer model, which outputs a point cloud based upon the sequence of tokens.
Systems and methods for automatic removal for selected training data from a dataset are provided. Systems and methods are provided for identifying portions of radar data that contain information that is not present in camera and/or lidar data. The identified portions of the radar data can be processed by a neural network to provide the additional information. The identified parts of the radar data can then be processed while the remaining parts of the radar data remain unprocessed. In various examples, features can be extracted from identified regions of the radar data using a neural network and fused with camera and/or lidar features. In some examples, the fused features can be used for object detection.
G06V 10/24 - Alignement, centrage, détection de l’orientation ou correction de l’image
G01S 7/41 - Détails des systèmes correspondant aux groupes , , de systèmes selon le groupe utilisant l'analyse du signal d'écho pour la caractérisation de la cibleSignature de cibleSurface équivalente de cible
G01S 13/42 - Mesure simultanée de la distance et d'autres coordonnées
G01S 13/86 - Combinaisons de systèmes radar avec des systèmes autres que radar, p. ex. sonar, chercheur de direction
G01S 13/931 - Radar ou systèmes analogues, spécialement adaptés pour des applications spécifiques pour prévenir les collisions de véhicules terrestres
G06V 10/40 - Extraction de caractéristiques d’images ou de vidéos
G06V 10/80 - Fusion, c.-à-d. combinaison des données de diverses sources au niveau du capteur, du prétraitement, de l’extraction des caractéristiques ou de la classification
G06V 10/82 - Dispositions pour la reconnaissance ou la compréhension d’images ou de vidéos utilisant la reconnaissance de formes ou l’apprentissage automatique utilisant les réseaux neuronaux
G06V 20/56 - Contexte ou environnement de l’image à l’extérieur d’un véhicule à partir de capteurs embarqués
95.
ASSOCIATION OF BOTTOM-UP KEYPOINTS BASED ON BOUNDING BOX EXTENTS
Systems and techniques are provided for associating vehicle keypoints with a bounding box. An example method includes receiving, by a machine learning model configured to perform object detection, sensor data from one or more sensors of an autonomous vehicle; identifying, based on the sensor data, at least one vehicle that is located within an environment of the autonomous vehicle; generating a bounding box corresponding to the at least one vehicle, wherein the bounding box includes one or more bounding box extents that are based on a type of the at least one vehicle; identifying, based on the sensor data, a plurality of vehicle keypoints each corresponding to a vehicle feature; and associating, based on at least one bounding box extent of the one or more bounding box extents, one or more vehicle keypoints from the plurality of vehicle keypoints with the bounding box.
Systems and techniques are provided for performing object detection with uncertainty predictions. An example method includes receiving, by a machine learning model configured to perform object detection, sensor data from one or more sensors of an autonomous vehicle; detecting, based on the sensor data, at least one object within an environment of the autonomous vehicle; determining, based on the sensor data, a plurality of object parameters associated with the at least one object; and determining, based on the sensor data, an uncertainty metric for each of the plurality of object parameters.
A radar sensor system comprises a first radar sensor and at least a second radar sensor and one or more processors configured to perform acts comprising transmitting a first signal from a first transmit antenna in a first radar sensor and transmitting a second signal from a second transmit antenna in a second radar sensor. The acts further comprise detecting an object at the first radar sensor and the second radar sensor and estimating vector velocity information vx and vy for the object. The acts also comprise generating a radar measurement vector z that comprises position information px and py for the object and incorporating the vector velocity information vx and vy into the radar measurement vector z. Additionally, the acts comprise iteratively performing a measurement update using the measurement vector z, with velocity information incorporated therein, and a linear Kalman filter until correct velocity values are determined.
Aspects of the subject technology relate to systems, methods, and computer-readable media for intelligently sampling data to train a model. Data for training a model can be accessed and separated into a first subset of data and a second subset of data. The model can be trained with the first subset of data to generate a trained model. Further, one or more errors associated with running the training model on the second subset of data can be identified. The second subset of data can be filtered to generate filtered data associated with the one or more errors. Additionally, the trained model can be further trained based on the filtered data to generate a refined trained model.
Disclosed are embodiments for facilitating purposeful stress testing of autonomous vehicle response time with simulation. In some aspects, an embodiment includes receiving a request to launch a simulation scenario on an autonomous vehicle (AV) that is to operate on a real-world test course; initiating a simulation derived from the simulation scenario using a simulation driver that is executing on the AV; engaging operation of the AV on the real-world test course; coordinating a simulated AV position in the simulation with a physical AV position of the AV on the real-world test course; combining virtual entities from the simulation with physical entities on the real-world test course into a list of tracked objects for the AV; and causing the AV to respond to the list of tracked objects including the virtual entities and the physical entities during the operation of the AV on the real-world test course.
G06F 30/20 - Optimisation, vérification ou simulation de l’objet conçu
B60W 50/04 - Détails des systèmes d'aide à la conduite des véhicules routiers qui ne sont pas liés à la commande d'un sous-ensemble particulier pour surveiller le fonctionnement du système d'aide à la conduite
B60W 60/00 - Systèmes d’aide à la conduite spécialement adaptés aux véhicules routiers autonomes
An ultrasonic detection system provides enhanced object detection capabilities. The ultrasonic detection system can generate complex ultrasonic waveforms based on excitation voltage waveforms constructed using high frequency components. In certain embodiments, an ultrasonic detection system may be capable of producing multiple different waveforms, e.g., different waveforms output by different transmitters, or multiple time-multiplexed waveforms output by the same transmitter. The ultrasonic detection system may be included in an autonomous vehicle (AV) for object detection in the environment of the AV.
brevets
