Automotive data processing system with efficient metadata generation and export - Patents.com

Abstract

The vehicle data processing system (20) includes a storage subsystem (44) and a processor (48). The storage subsystem is located within the vehicle and configured to store at least data (45) produced by one or more data sources (24, 28, 30, 32, 34) of the vehicle. The processor is installed in the vehicle and configured to apply at least one model to data stored in or in the process of being stored within the storage subsystem, the model identifying one or more specified features of interest in the data, generate metadata (46) that tags occurrences of the specified features of interest in the stored data, and export at least a portion of the metadata to an external system external to the vehicle.

Description

[Technical field]

[0001] [CROSS REFERENCE TO RELATED APPLICATIONS] This application claims the benefit of U.S. Provisional Patent Application No. 62 / 948,027, filed December 13, 2019, the disclosure of which is incorporated herein by reference.

[0002] The present disclosure relates generally to automotive data processing, and more particularly to methods and systems for in-vehicle metadata generation and export. [Background technology]

[0003] Autonomous and other vehicles generate and process large amounts of data, sometimes on the order of terabytes, during the course of a day's operation. Sources of data in autonomous vehicles include, for example, cameras and other sensors, advanced driver assistance systems (ADAS), telematics control units (TCUs), infotainment systems, and various electronic control units (ECUs). Traditionally, large amounts of data are periodically uploaded to a remote processor.

[0004] The above discussion is presented as a summary of the relevant art in the field and should not be construed as an admission that any information it contains constitutes prior art to the present patent application. Summary of the Invention

[0005] Embodiments described herein provide an automobile data processing system including a storage subsystem and a processor, the storage subsystem disposed within a vehicle and configured to store at least data produced by one or more data sources of the vehicle, the processor installed in the vehicle and configured to apply to the data stored within the storage subsystem or en route to the storage subsystem at least one model that identifies one or more specified features of interest in the data to generate metadata that tags occurrences of the specified features of interest in the stored data, and export at least a portion of the metadata to an external system external to the vehicle.

[0006] In some embodiments, the processor is configured to identify occurrences of the same feature of interest in the data produced by two or more different ones of the data sources and, when generating the metadata, tag the data to link the occurrences of the same feature of interest that appear in the data produced by the two or more different ones of the data sources to one another. In one embodiment, the data produced by the two or more ones of the data sources is the same type of data. In another embodiment, the data produced by the two or more ones of the data sources is different types of data. In yet another embodiment, the two or more ones of the data sources are located at different locations within the vehicle.

[0007] In some embodiments, the processor is further configured to receive a request from the external system for one or more selected portions of the data related to the feature of interest in response to the exported metadata, and to provide at least the selected portions, but less than all of the data, to the external system. In one exemplary embodiment, the processor is configured to provide two or more selected portions of the data to the external system, the two or more selected portions being produced by two or more different ones of the data sources and associated with the same feature of interest.

[0008] In one disclosed embodiment, the processor is configured to select the portions of the metadata to export according to a selection criterion, hi one embodiment, the processor is configured to export the metadata without simultaneously exporting data corresponding to the metadata.

[0009] In one exemplary embodiment, the storage subsystem is configured to store at least data produced by multiple independent data sources within the vehicle. In other embodiments, the storage subsystem is configured to store at least data produced by multiple data sources installed at different locations within the vehicle. In one embodiment, the storage subsystem is configured to store two or more different types of the data.

[0010] In some embodiments, the processor is configured to receive the model from the external system. In one embodiment, the processor is configured to export the metadata while the vehicle is connected to a docking station. In other embodiments, the processor is configured to export the metadata regardless of whether the vehicle is connected to a docking station.

[0011] In some embodiments, the model includes an artificial intelligence (AI) inference model, and the processor is configured to generate the metadata by applying the AI ​​inference model to the data. In one embodiment, the processor is configured to receive a pre-trained AI inference model from the external system and update training of the AI ​​inference model within the vehicle.

[0012] In some embodiments, the storage subsystem includes a centralized storage device configured to store the data produced by a plurality of the data sources, hi one embodiment, the centralized storage device is configured to store the data produced by two or more of the data sources, each of the data sources being of a different type.

[0013] In one embodiment, the processor is further configured to apply a second model to the data, different from the first model, the second model identifying one or more second features of interest in the data, add a second occurrence of the second feature of interest to the metadata according to the second model, and provide one or more of: (i) metadata relating to only the occurrence of the feature of interest, (ii) metadata relating to only the second occurrence of the second feature of interest, and (iii) metadata relating to both the occurrence and the second occurrence to the external system. In one embodiment, the processor is configured to store the metadata in a structured form within the storage subsystem.

[0014] According to embodiments described herein, there is further provided a method of processing automobile data, comprising storing in a storage subsystem located in a vehicle at least data produced by one or more data sources of the vehicle, applying, using a processor installed in the vehicle, at least one model that identifies one or more specified features of interest in the data to the data stored in or en route to the storage subsystem to generate metadata that tags occurrences of the specified features of interest in the stored data, and exporting at least a portion of the metadata to an external system external to the vehicle.

[0015] According to embodiments described herein, there is further provided an automobile data processing system including a plurality of data sources and a processor, the data sources being installed in a vehicle and configured to generate data, the processor being installed in the vehicle and configured to collect the data generated by the data sources, locally store the data in a storage device located in the vehicle without exporting the data outside the vehicle, identify one or more portions of interest in the locally stored data, and upload at least the identified portions of interest, but less than all of the locally stored data, to an external system while the vehicle is connected to a docking station coupled to an external system located remotely from the docking station.

[0016] In some embodiments, the processor is configured to collect and store the data and identify the portion of interest regardless of whether the vehicle is connected to a docking station. In one embodiment, the processor is configured to collect the data from two or more different ones of the data sources that produce the same type of data. In other embodiments, the processor is configured to collect the data from two or more different ones of the data sources that produce different types of data.

[0017] In yet another embodiment, the processor is configured to collect the data from two or more different ones of the data sources located at different locations within the vehicle. In one disclosed embodiment, the processor is configured to identify and upload two or more parts of interest produced by two or more different ones of the data sources and associated with a same feature of interest.

[0018] Also provided in accordance with embodiments described herein is a method for automobile data processing, including generating data by a plurality of data sources installed in a vehicle, using a processor installed in the vehicle to collect the data generated by the data sources, locally store the data in a storage device located in the vehicle without exporting the data outside the vehicle, identify one or more interesting portions of the locally stored data, and upload at least the identified interesting portions, but less than all, of the locally stored data to the external system while the vehicle is connected to a docking station coupled to an external system located remotely from the docking station.

[0019] In accordance with embodiments described herein, there is further provided an automobile data processing system including a plurality of data sources, a packet network, and a processor, the data sources being distributed at different locations within a vehicle and configured to generate data, the packet network being disposed within the vehicle and configured to transfer the data from the plurality of data sources to a central storage location within the vehicle, the processor being installed in the vehicle and configured to generate metadata corresponding to the data transferred from the plurality of data sources and stored at the central storage location, and transfer one or more selected portions of the sensor data to an external system external to the vehicle, the selected portions being selected based on the metadata corresponding to the data transferred from the plurality of data sources and stored at the central storage location.

[0020] In one embodiment, the plurality of data sources includes two or more different data sources configured to produce different types of data. In other embodiments, the data produced by the two or more of the data sources are different types of data. In one disclosed embodiment, the processor is configured to select the one or more portions of the sensor data at least in part in response to an indication of one or more features of interest received from the external system.

[0021] In one exemplary embodiment, the processor is configured to generate the metadata by applying an artificial intelligence (AI) inference model to the data stored in the central storage location. In one embodiment, in generating the metadata, the processor is configured to tag one or more occurrences of one or more features of interest in the data. In another embodiment, in generating the metadata, the processor is configured to tag occurrences of the same feature of interest in two or more different types of the data or in data generated by two or more different data sources.

[0022] According to embodiments described herein, there is further provided a method for automobile data processing, comprising: generating data by a plurality of data sources distributed at different locations within a vehicle; transferring the data from the plurality of data sources to a central storage location within the vehicle via a packet network located within the vehicle; generating metadata corresponding to the data transferred from the plurality of data sources and stored in the central storage location using a processor installed in the vehicle; transferring one or more selected portions of the sensor data to an external system external to the vehicle, the selected portions being selected based on the metadata corresponding to the data transferred from the plurality of data sources and stored in the central storage location.

[0023] Also provided in accordance with embodiments described herein is an automobile data analysis system including an interface for communicating with a vehicle and a computer configured to define a model that identifies one or more specified features in data produced by one or more data sources of the vehicle, provide the model to a processor installed in the vehicle, receive metadata from the processor in the vehicle generated by the processor using the model, the metadata tagging occurrences of the specified features in the data, and analyze the received metadata.

[0024] In one embodiment, the interface is configured to receive the metadata while the vehicle is connected to a docking station, while in other embodiments, the interface is configured to receive the metadata from the vehicle regardless of whether the vehicle is connected to a docking station.

[0025] In some embodiments, the model includes an artificial intelligence (AI) inference model, and the computer is configured to train the AI ​​inference model and provide the trained AI inference model to the processor in the vehicle. In one disclosed embodiment, the computer is configured to send a request for one or more selected portions of the stored data to the processor in the vehicle in response to the received metadata, receive the requested one or more selected portions from the processor, and analyze the one or more selected portions of the data.

[0026] In other embodiments, the computer is configured to request two or more selected portions of the stored data, the two or more selected portions being produced by two or more different ones of the data sources. In yet other embodiments, the computer is configured to request two or more selected portions of the stored data, the two or more selected portions being associated with a same feature of interest.

[0027] According to embodiments described herein, there is further provided a method of analyzing automobile data, comprising defining, in a computer external to a vehicle, a model that identifies one or more specified features in data produced by one or more data sources of the vehicle, and providing the model to a processor installed in the vehicle. Metadata is received in the computer external to the vehicle from the processor in the vehicle, the metadata being generated by the processor using the model and tagging occurrences of the specified features in the data. The received metadata is analyzed.

[0028] The present disclosure will become more fully understood from the following detailed description of the embodiments thereof taken in conjunction with the drawings, in which: [Brief description of the drawings]

[0029] [Figure 1] FIG. 1 is a block diagram that illustrates a schematic of an automobile data processing system according to embodiments described herein.

[0030] [Diagram 2] 2 is a flow diagram that generally illustrates a method for generating and processing metadata within the automotive system of FIG. 1 according to embodiments described herein. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0031] Modern vehicles generate vast amounts of data of various types originating from various data sources. An autonomous vehicle may generate, for example, one or more video feeds from a video camera, various types of sensor data streamed from sensors located at various locations within the vehicle, radar and lidar information, driver behavior information, information from infotainment subsystems and user applications, and location and navigation information such as GPS data and GPS maps. In a typical day of driving, an autonomous vehicle may generate on the order of four terabytes of data.

[0032] Data generated within vehicles in operation can provide extremely valuable information if analyzed correctly: GPS data and car driving data can be used to update commute prediction models, camera footage can be used to optimize autonomous driving performance and safety procedures for autonomous vehicles, and driving behavior information can be used by insurance companies to determine driver premiums, just to name a few.

[0033] Typically, the computational power required to perform such analyses far exceeds the capabilities of current state-of-the-art or available in-vehicle computing resources. Thus, in various embodiments, at least some of the raw data is exported to an external system, such as, for example, a cloud-based data processing system, in order to effectively analyze the data. On the other hand, due to time, bandwidth and power consumption limitations, it is not possible to transfer all of the raw data from the vehicle to the external system. These constraints limit the effectiveness of analyzing data generated within the vehicle and extracting valuable and actionable information therefrom.

[0034] In view of the above, the embodiments described herein provide improved methods and systems for processing, communicating and analyzing data generated within a vehicle.

[0035] In some disclosed embodiments, an automobile data processing system is deployed within a vehicle. The automobile data processing system includes a storage subsystem coupled to a processor. During operation of the vehicle, the storage subsystem stores raw data produced by various data sources. The processor pre-processes the raw data by identifying designated features of interest and generates metadata that tags occurrences of the designated features of interest in the raw data. The metadata, or at least selected portions thereof, is exported from the vehicle to an external system.

[0036] The external system typically analyzes the metadata to identify specific portions of the raw data that are of interest and retrieves these specific portions from the vehicle for analysis. In one embodiment, the automobile data processing system is configured to communicate with the external system while the vehicle is connected to the docking station. In this embodiment, the processor stores the metadata locally in the storage subsystem on an ongoing basis. When the vehicle is docked, i.e., parked and connected, to the docking station, the processor transmits the accumulated metadata to the external system.

[0037] In some embodiments, an on-board processor identifies features of interest and generates metadata by applying a pre-trained artificial intelligence (AI) inference model to the raw data. The AI ​​model may be provided, for example, by an external system. In some embodiments, the processor applies several different AI models, for example, one each for identifying different features of interest or each for identifying the same feature of interest in different types of raw data. Some example implementations and illustrative use cases of the disclosed technology are described below.

[0038] In some embodiments, the same feature of interest (e.g., an object or event) may appear in two or more different raw data streams produced by two or more data sources (e.g., sensors of the same type but located at different locations in the vehicle, or sensors of different types located at the same or different locations in the vehicle). In some embodiments, an on-board processor identifies features of interest (e.g., objects or events) in the different raw data streams and generates metadata that tags portions of the raw data that are common to and shared among occurrences of the same feature of interest in the different streams. Further aspects of generating and processing metadata for different media objects are addressed in U.S. Patent Application Publication No. 2020 / 0042548, entitled "Metadata Generation for Multiple Object Types," the disclosure of which is incorporated herein by reference.

[0039] In general, metadata is orders of magnitude smaller than raw data and has a simpler and structured format. Therefore, it is possible to transmit the metadata to an external system for a short period of time using a reasonable bandwidth. The specific portion of raw data that the external system then requests based on the metadata is also much smaller than the total amount of raw data. Thus, the disclosed technology (i) overcomes the communication bottleneck between the in-vehicle system and the external system, and (ii) provides an optimal "division of labor" that does not impair the quality of data analysis.

[0040] 1 is a block diagram that illustrates a schematic of an automobile data processing system 20 according to an embodiment described herein. The system 20 is installed in a vehicle and includes various sensors 24, a number of electronic control units (ECUs) 32, an advanced driver assistance system (ADAS) 28, an infotainment system 30, and a central computer 34.

[0041] The sensors 24 may include, for example, video cameras, speed sensors, accelerometers, audio sensors, infrared sensors, radar sensors, lidar sensors, ultrasonic sensors, distance measuring devices or other proximity sensors, or any other suitable sensor type. In this example, each ECU 32 (sometimes referred to as a "zone ECU") is connected to sensors installed in a respective zone of the vehicle. Generally, each ECU 32 controls its respective sensor 24 and collects data from those sensors. In one embodiment, one or more of the sensors 24, such as image sensors, are connected directly to the ADAS 28 without going through the ECU 32.

[0042] In the present context, any element in system 20 that produces data (including, for example, sensors 24, ECU 32, ADAS 28, infotainment system 30, and central computer 34) is considered a "data source." ECU 32, ADAS 28, infotainment system 30, and central computer 34 are considered examples of "electronic subsystems" of the vehicle. It will be appreciated that certain system components can act as both a data source and an electronic subsystem.

[0043] In some embodiments, the various electronic subsystems of system 20 are deployed in different locations within the vehicle and communicate over a packet network installed in the vehicle. In this example, the packet network comprises an Ethernet network, although other suitable network protocols may be used. The network comprises a number of Ethernet links 36 and one or more Ethernet switches 40. In various embodiments, the bit rate used within the network may be 10 Gbit / s (10 Gbps) according to IEEE 802.3ch, 1000 Mbps according to IEEE 802.3bp, 100 Mbps according to IEEE 802.3bw, 10 Mbps according to IEEE 802.3cg (10Base-T1s), or any other suitable bit rate. Links 36 may comprise, for example, twisted pair copper links, or any other type of link suitable for Ethernet communication.

[0044] 1, system 20 further includes a centralized storage device 44 (also referred to as a “storage subsystem”) and a processor 48. Aspects of centralized storage within automotive networks are addressed, for example, in U.S. patent application Ser. No. 17 / 094,844, filed Nov. 11, 2020, entitled “Automotive Network with Centralized Storage,” the disclosure of which is incorporated herein by reference.

[0045] The storage device 44 stores raw data 45 provided by various data sources (e.g., sensors and / or electronic subsystems) within the vehicle. In some embodiments, the processor 48 identifies designated features of interest within the raw data 45, generates metadata 46 that tags occurrences of the designated features of interest in the raw data, and stores the metadata 46 on the storage device 45. In one embodiment, the processor 48 identifies the features of interest by applying a computational artificial intelligence (AI) engine to the raw data 45. The processor 48 exports at least a portion of the metadata to an external system outside the vehicle, for example, a cloud-based data processing system or other suitable computer outside the vehicle.

[0046] In one embodiment, the centralized storage device 44 comprises a solid-state drive (SSD) directly connected to the packet network installed in the vehicle. As can be seen, in this example, the centralized storage device 44 comprises two interfaces connected to two different ports of two different switches 40 via respective links 36. In this manner, a degree of redundancy is provided, since failure of a single interface (or corresponding link) does not isolate the centralized storage device 44 from the network.

[0047] In this example, processor 48 and storage device 44 communicate with each other via a high-speed computer bus, such as a Peripheral Component Interconnect Express (PCIe) bus. Alternatively, in some embodiments, processor 48 may be independently connected to a packet network, in which case processor 48 and storage device 44 communicate with each other via this network. Processor 48 is configured to communicate with a docking station to which the vehicle is connected using data link 52, or any other suitable interface for transferring data (not shown).

[0048] Typically, data sources (sensors and / or electronic subsystems) send raw data 45 over a network to a centralized storage device 44. In various embodiments, the data sources and centralized storage device 44 may communicate using any suitable protocol, such as the Non-Volatile Memory Express over Fabric (NVMe-oF) protocol or NVMe-over-TCP. Further aspects of centralized storage that may be applied within system 20 are described in U.S. patent application Ser. No. 17 / 094,844, filed Nov. 11, 2020, and entitled “Automotive Network with Centralized Storage,” the disclosure of which is incorporated herein by reference.

[0049] The configuration of communications system 20 and its components (such as various devices and electronic subsystems and / or centralized storage device 44) shown in FIG. 1 is an exemplary configuration shown for clarity only. In alternative embodiments, any other suitable configuration may be used. For example, system 20 may include other suitable types of electronic subsystems and / or devices laid out and connected in any other suitable manner. As another example, the packet network (including links 36 and switches 40) may have any other suitable topology.

[0050] Different elements of system 20 and its components may be implemented using dedicated hardware or firmware (such as using hardwired or programmable logic), for example in an application specific integrated circuit (ASIC) or field programmable gate array (FPGA). Additionally or alternatively, some functions of the components of system 20, such as some functions of ECU 32, storage device 44 and / or processor 48, may be implemented in software and / or using a combination of hardware and software elements. Elements that are not essential for an understanding of the disclosed technology have been omitted from the figures for clarity.

[0051] In some embodiments, some functions of ECU 32, storage device 44 and / or processor 48 may be implemented in one or more programmable processors that are programmed in software to perform the functions described herein. This software may be downloaded to either processor in electronic form over a network, for example, and / or may alternatively be provided on and / or stored on a non-transitory tangible medium, such as a magnetic, optical or electronic memory.

[0052] As described above, various data sources within the vehicle (e.g., sensors and electronic subsystems) produce raw data 45 during ongoing operation of the vehicle and transmit the data for storage at centralized storage device 44. Raw data 45 may include, for example, video data, audio data, sensor data, radar and / or lidar data, driver behavior data, data from infotainment subsystems and / or user applications, location and / or navigation data, such as GPS data and GPS maps, and / or any other suitable type of data.

[0053] It will be appreciated that at least some of the raw data (e.g., video) may be unstructured. The raw data is typically of multiple different types (modalities) and typically originates from multiple independent data sources of different types installed at multiple different locations within the vehicle. In some embodiments, multiple raw data streams are generated by multiple sensors of the same type (e.g., multiple video cameras or multiple lidars), each of which is located at a different location within the vehicle. In other embodiments, multiple instances of the same type of sensor are located at the same location within the vehicle to provide redundant streams of raw data of the same type.

[0054] Processor 48 scans raw data 45 to identify one or more specified features of interest, generates a database of metadata 46 that tags occurrences of the specified features of interest in the raw data, and stores metadata 46 in storage device 45. In various embodiments, processor 48 may search for any suitable features of interest, such as, for example: -Objects or environmental conditions near the vehicle that are visible in the video or detectable in the sensor data (e.g., obstacles, traffic signs, other vehicles or pedestrians approaching the vehicle). - Events related to driver behavior (e.g. slowing down or accelerating the vehicle after reacting to a traffic light, steering sharply to avoid an accident or obstacle, or responding to a traffic sign). -Events related to the vehicle's behavior on the road (e.g. the effects of autonomous driving). - Events related to the presence or behavior of nearby vehicles, for example heavy or light traffic scenarios. -Events related to driver intervention in the autonomous operation of the vehicle. -Communication with external elements (e.g. other vehicles, lighting, road sensors). -Mechanical symptoms within the vehicle (e.g., change in responsiveness or noise due to loss of tire pressure, wear on one or more brake pads, loss of visibility due to burnt out headlights or other lights, reduced connectivity due to issues with the in-vehicle network).

[0055] The above list of features of interest is by way of example only, and processor 48 may additionally or alternatively identify any other suitable features that may be of interest for analysis.

[0056] As noted above, in some cases, a particular feature of interest (e.g., an object or event) occurs in the raw data of multiple data sources. The multiple data sources (and thus raw data streams) may be of different types (e.g., video, audio, lidar, radar, etc.) or multiple instances of the same type of raw data (e.g., video captured at different locations in the vehicle). For example, a road accident or a mechanical malfunction or failure may be seen in a video feed, heard in audio data, detected by various sensors, and / or reported by one or more electronic subsystems. The correlation between such occurrences of a given feature of interest in the raw data of different data sources is itself valuable. Thus, in some embodiments, processor 48 tags occurrences of such features of interest (e.g., objects or events) with the same tag or otherwise links occurrences in the data of different sensors to one another.

[0057] In various embodiments, processor 48 can use a variety of techniques for identifying specified features of interest in raw data 45. In some embodiments, processor 48 executes a computational artificial intelligence (AI) engine configured to compile and apply pre-trained AI inference models to identify features of interest in raw data 45. In some embodiments, processor 48 applies several different AI models to the raw data, e.g., each model for identifying a different feature of interest, or each model for identifying the same feature of interest in different types of raw data.

[0058] In one embodiment, processor 48 receives one or more pre-trained artificial intelligence (AI) models from an external system, for example while the vehicle is connected to a docking station, via data link 52. During ongoing operation of the vehicle, processor 48 applies the one or more pre-trained AI models to raw data 45 provided by various data sources.

[0059] Processor 48 uses one or more AI models to identify occurrences of features of interest in raw data 45 and generate metadata 46 that tags these occurrences. Processor 48 can apply the AI ​​models to raw data already stored in storage device 44 and / or to raw data en route to storage device 44. Metadata 46 is temporarily stored in storage device 44. At an appropriate time, for example when the vehicle is connected to a docking station, processor 48 transmits metadata 46 calculated within the vehicle to an external system, generally without transferring all of the raw data.

[0060] It should be noted that since the data size of the metadata 46 is certainly relatively small compared to the raw data, in some embodiments, it is not required that the vehicle use its own connection to the docking station to transmit the metadata to the external system. In one embodiment, the processor 48 can transmit the metadata 46 to the external system during vehicle operation, regardless of whether the vehicle is connected to the docking station, via a suitable wireless link, such as, for example, a cellular link, a mobile Wi-Fi® connectivity coupling, or other suitable wireless link. Similarly, in one embodiment, the external system can send the AI ​​model to the processor 48, for example, via a suitable wireless link, regardless of whether the vehicle is connected to the docking station. In general, communication via the docking station (e.g., receiving the AI ​​model and exporting requested portions of the metadata and / or raw data) is considered offline processing.

[0061] In some embodiments, processor 48 transmits all of metadata 46 to the external system. In other embodiments, processor 48 uses selection criteria to select only a portion of metadata 46 and transmits only the selected portion of the metadata to the external system.

[0062] The disclosed technology can be used to analyze a wide variety of features in a variety of use cases. For example, consider a use case in which an external system analyzes the behavior of an autonomous vehicle when an animal appears on or around the road. To perform this type of analysis, in one embodiment, the external system trains an AI model to recognize animals in raw video data. Processor 48 receives the pre-trained AI model, compiles it, and applies it to the video camera feeds provided by the various cameras in the vehicle. Based on the identified occurrences of the features of interest (appearance of the animal in the video), processor 48 generates metadata that tags the video frames in which the animal appears. The tagged frames can be identified in the video streams from two or more video sensors located at different locations in the vehicle. Processor 48 uploads the metadata to the external system for initial analysis.

[0063] In one embodiment, the external system can then request the portion of raw data 45 of interest that pertains to the identified occurrence (e.g., a snippet of the footage beginning shortly before the animal's appearance). Note that the requested portion of raw data of interest does not necessarily have to be the exact portion where the occurrence was identified. For example, the external system can record the times when animals were identified in the footage and request the sensor data corresponding to these times. In one embodiment, processor 48 uploads the portion of raw data of interest (e.g., along with metadata) without being requested by the external system.

[0064] As another example, consider a use case in which an external system analyzes the behavior of an autonomous vehicle in rainy conditions when pedestrians are present. To perform this type of analysis, in one embodiment, the external system provides processor 48 with multiple pre-trained AI models (e.g., a model for recognizing rainy scenes in video, other models for recognizing pedestrians in video, and possibly other models for recognizing rainy conditions in the output of other types of sensors, such as environmental condition sensors). Using these models, processor 48 generates metadata that tags time intervals in the video where the scene is raining, time intervals in the video that show pedestrians, and possibly such time intervals in other sensor raw data. The external system can then find the intersection between the various types of intervals (i.e., identify time intervals of rainy conditions that include pedestrians) and request sensor and / or subsystem data that correspond to these time intervals.

[0065] The above two examples are highly simplified and in no way limiting, and are provided only to demonstrate the effectiveness of the disclosed technology. Other examples may include, for example, more data sources, each of different types.

[0066] 2 is a flow diagram that generally illustrates a method for generating and processing metadata according to embodiments described herein. The left side of the diagram illustrates actions performed by an external system (e.g., a cloud-based data processing system). The right side of the diagram illustrates actions performed by the system 20 in the vehicle.

[0067] In a model training operation 60, the external system trains an AI inference model to identify a particular feature of interest in a particular type of raw data 45 (e.g., to identify people in video data, to identify mechanical faults in audio or sensor data, etc.). The trained model is provided to the processor 48, for example (but not necessarily) while the vehicle is connected to a docking station. In a compilation operation 64, the processor 48 compiles the AI ​​model for execution on the processor's computational AI engine. In one embodiment, the processor 48 can update the training of the AI ​​model to take into account local inputs (e.g., perform additional training using data obtained in the particular vehicle). Such updates can, for example, help to better adapt to the area in which the particular vehicle typically operates. This kind of additional training can be performed at any stage, for example while the vehicle is connected to the docking station and / or while driving.

[0068] In a data collection operation 68, various data sources within the vehicle generate raw data 45 and transmit the raw data for storage on the storage device 44. In an identification operation 72, the processor 48 identifies occurrences of features of interest in the raw data 45 by applying an AI model.

[0069] In a metadata generation operation 76, processor 48 generates metadata 46 that tags the identified occurrences. Each tag in metadata 46 generally indicates the identified feature of interest and the location of the occurrence in the raw data (e.g., a frame number in a video or other location depending on the type of raw data). Processor 48 can apply the AI ​​model to raw data already stored in storage device 44 and / or to raw data in the process of being stored to storage device 44.

[0070] As described above, metadata generation can be applied to multiple raw data sets and can involve matching tags of the same feature of interest (e.g., object or event) that appear in different raw data sets provided by different sensors of the same or different types.

[0071] At a metadata store operation 80, the processor 48 stores the metadata 46 in the storage device 44. At a notification operation 84, the processor 48 notifies the external system that the metadata 46 is available. At a metadata retrieve operation 88, the external system fetches the metadata 46 from the vehicle's storage device 44. In various embodiments, the metadata 46 may be "pulled" from the vehicle by the external system or "pumped" to the external system at the vehicle's initiative. For example, the processor 48 may push the metadata to the external system periodically or when the vehicle is connected to a docking station.

[0072] In an initial analysis operation 92, the external system performs a suitable initial analysis on the metadata. The initial analysis typically involves determining one or more interesting portions of the raw data 45 related to features of interest based on the metadata. In a data retrieval operation 96, the external system fetches the interesting portions of the raw data 45 from the vehicle's storage device 44. The external system then analyzes the interesting portions of the raw data in an analysis operation 100.

[0073] In one embodiment, the above process can be repeated with a different AI model that the processor 48 remaps and applies to the computational AI engine to tag different features of interest (or the same features of interest in different types of raw data). The processor 48 can add new metadata to the existing metadata database or generate a separate database of metadata 46 to be stored in the storage device 44.

[0074] Although the embodiments described herein primarily deal with raw data obtained within automotive systems, the methods and systems described herein may also be used in other applications, such as, for example, in processing various other types of unstructured raw data, such as video library analysis and audio readout analysis.

[0075] It is noted that the above embodiments are cited as examples, and the present invention is not limited to those specifically shown and described above.On the contrary, the scope of the present invention includes both combinations and subcombinations of the various features described above, as well as variations and modifications thereof, which would occur to a person skilled in the art upon reading the above description, and which are not disclosed in the prior art.Documents incorporated by reference into this patent application are considered to be an integral part of this application, with the exception that to the extent that the terms defined in these incorporated documents are inconsistent with the definitions expressly or impliedly made herein, only the definitions within this specification should be considered. [Other possible items] [Item 1] a storage subsystem disposed within the vehicle and configured to store at least data produced by one or more data sources of the vehicle; A processor installed in a vehicle, applying at least one model to the data stored in or en route to the storage subsystem that identifies one or more specified features of interest in the data to generate metadata that tag occurrences of the specified features of interest in the stored data; a processor configured to export at least a portion of the metadata to an external system external to the vehicle; An automobile data processing system comprising: [Item 2] 2. The automobile data processing system of claim 1, wherein the processor is configured to identify occurrences of the same feature of interest in the data produced by two or more different data sources of the plurality of data sources and, when generating the metadata, to tag the data to link together the occurrences of the same feature of interest that appear in the data produced by the two or more different data sources of the plurality of data sources. [Item 3] 3. The automobile data processing system of claim 2, wherein the data produced by the two or more data sources of the plurality of data sources is the same type of data. [Item 4] 3. The automobile data processing system of claim 2, wherein the data produced by the two or more data sources of the plurality of data sources are different types of data. [Item 5] 3. The automobile data processing system of claim 2, wherein the two or more data sources of the plurality of data sources are located at different locations within the vehicle. [Item 6] The processor, receiving a request from the external system for one or more selected portions of the data related to the feature of interest in response to the exported metadata; and further configured to provide at least the selected portion of the data, but less than all of the data, to the external system. 6. An automobile data processing system according to any one of items 1 to 5. [Item 7] 7. The automobile data processing system of claim 6, wherein the processor is configured to provide two or more selected portions of the data to the external system, the two or more selected portions being produced by two or more different data sources of the plurality of data sources and associated with the same feature of interest. [Item 8] 6. The automobile data processing system of any one of claims 1 to 5, wherein the processor is configured to select the portion of the metadata to export according to a selection criterion. [Item 9] 6. The motor vehicle data processing system of claim 1, wherein the processor is configured to export the metadata without simultaneously exporting data corresponding to the metadata. [Item 10] 6. The automobile data processing system of any one of claims 1 to 5, wherein the storage subsystem is configured to store at least data produced by a plurality of independent data sources within the vehicle. [Item 11] 6. The automobile data processing system of any one of claims 1 to 5, wherein the storage subsystem is configured to store at least data produced by a plurality of data sources installed at different locations within the vehicle. [Item 12] 6. The automobile data processing system of any one of claims 1 to 5, wherein the storage subsystem is configured to store two or more different types of the data. [Item 13] 6. The vehicle data processing system of any one of claims 1 to 5, wherein the processor is configured to receive the model from the external system. [Item 14] 6. The automobile data processing system of claim 1, wherein the processor is configured to export the metadata while the vehicle is connected to a docking station. [Item 15] 6. The automobile data processing system of any one of claims 1 to 5, wherein the processor is configured to export the metadata regardless of whether the vehicle is connected to a docking station. [Item 16] 6. The automobile data processing system of claim 1, wherein the model includes an artificial intelligence (AI) inference model, and the processor is configured to generate the metadata by applying the AI ​​inference model to the data. [Item 17] Item 17. The automobile data processing system of item 16, wherein the processor is configured to receive a pre-trained AI inference model from the external system and update training of the AI ​​inference model within the vehicle. [Item 18] 6. The automobile data processing system of claim 1, wherein the storage subsystem includes a centralized storage device configured to store the data produced by the multiple data sources. [Item 19] 20. The automobile data processing system of claim 18, wherein the centralized storage device is configured to store the data produced by two or more data sources of different types among the plurality of data sources. [Item 20] The processor, applying a second model to the data, the second model being different from the first model, the second model identifying one or more second features of interest in the data; adding a second occurrence of the second feature of interest to the metadata according to the second model; and and (iii) providing to the external system one or more of: (i) metadata regarding only the occurrence of the feature of interest; (ii) metadata regarding only the second occurrence of the second feature of interest; and (iii) metadata regarding both the occurrence and the second occurrence. 6. An automobile data processing system according to any one of items 1 to 5. [Item 21] 6. The automobile data processing system of claim 1, wherein the processor is configured to store the metadata in a structured form within the storage subsystem. [Item 22] storing, in a storage subsystem located within the vehicle, at least data produced by one or more data sources of the vehicle; Using a processor installed in the vehicle, applying at least one model to the data stored in or en route to the storage subsystem that identifies one or more specified features of interest in the data to generate metadata that tag occurrences of the specified features of interest in the stored data; and exporting at least a portion of the metadata to an external system external to the vehicle. and using the An automobile data processing method comprising: [Item 23] 23. The method for processing automobile data according to claim 22, wherein the step of applying the model comprises identifying occurrences of the same feature of interest in the data produced by two or more different data sources of the plurality of data sources, and the step of generating the metadata comprises tagging the data to link together the occurrences of the same feature of interest that appear in the data produced by the two or more different data sources of the plurality of data sources. [Item 24] 24. The method for processing vehicle data according to claim 23, wherein identifying the occurrence of the same feature of interest comprises identifying the occurrence in data provided by two or more of the plurality of data sources, the two or more data sources being located at different positions within the vehicle. [Item 25] receiving a request from the external system in response to the exported metadata for one or more selected portions of the data related to the feature of interest; providing at least the selected portion of the data, but less than all of the data, to the external system; 25. The automobile data processing method according to any one of claims 22 to 24, further comprising: [Item 26] 26. The method of processing vehicle data of claim 25, wherein providing the selected portions comprises providing two or more selected portions of the data to the external system, the two or more selected portions being produced by two or more different data sources of the plurality of data sources and associated with the same feature of interest. [Item 27] 25. The method for processing vehicle data according to any one of claims 22 to 24, wherein the step of exporting the metadata comprises the steps of exporting the metadata separately from data corresponding to the metadata, selecting one or more interesting parts of the data based on the metadata, and exporting the selected one or more interesting parts of the data when the vehicle is connected to a docking station. [Item 28] A plurality of data sources installed in the vehicle and configured to generate data; a processor installed in the vehicle, collecting the data generated by the plurality of data sources and storing the data locally in a storage device located within the vehicle without exporting the data outside the vehicle; Identifying one or more portions of interest within the locally stored data; a docking station configured to upload at least the identified portion of interest, but less than all of the locally stored data, to an external system while the vehicle is connected to the docking station coupled to the external system located remotely from the docking station; Processor and An automobile data processing system comprising: [Item 29] 30. The automobile data processing system of claim 28, wherein the processor is configured to collect and store the data and identify the portion of interest regardless of whether the vehicle is connected to a docking station. [Item 30] 30. The automobile data processing system of claim 28, wherein the processor is configured to collect the data from two or more different data sources among the plurality of data sources that produce the same type of data. [Item 31] 30. The automobile data processing system of claim 28, wherein the processor is configured to collect the data from two or more different data sources of the plurality of data sources, each producing a different type of data. [Item 32] 30. The automobile data processing system of claim 28, wherein the processor is configured to collect the data from two or more different data sources of the plurality of data sources, the different data sources being located at different positions within the vehicle. [Item 33] 33. The vehicle data processing system of any one of items 28 to 32, wherein the processor is configured to identify and upload two or more parts of interest that are produced by two or more different data sources of the plurality of data sources and that are associated with the same feature of interest. [Item 34] generating data from a plurality of data sources installed in the vehicle; Using a processor installed in the vehicle, collecting the data generated by the data sources and storing the data locally in a storage device located within the vehicle without exporting the data outside of the vehicle; identifying one or more portions of the locally stored data that are of interest; and uploading at least the identified portion of interest, but less than all of the locally stored data, to an external system while the vehicle is connected to a docking station coupled to the external system located remotely from the docking station. and using the An automobile data processing method comprising: [Item 35] 35. The method for processing automobile data according to claim 34, wherein the steps of collecting and storing the data and identifying the part of interest are performed regardless of whether the vehicle is connected to a docking station. [Item 36] Item 35. The automobile data processing method of item 34, wherein the step of collecting data comprises a step of collecting the data from two or more different data sources among the plurality of data sources that produce the same type of data. [Item 37] Item 35. The automobile data processing method of item 34, wherein the step of collecting data comprises a step of collecting the data from two or more different data sources of the plurality of data sources, each of the data sources producing a different type of data. [Item 38] Item 35. The automobile data processing method of item 34, wherein the step of collecting data comprises a step of collecting the data from two or more different data sources among the plurality of data sources, the different data sources being located at different positions within the vehicle. [Item 39] 39. The method for processing vehicle data according to any one of claims 34 to 38, wherein the steps of identifying and uploading the parts of interest comprise identifying and uploading two or more parts of interest, the two or more parts of interest being produced by two or more different data sources of the plurality of data sources and associated with the same feature of interest. [Item 40] A plurality of data sources distributed at different locations within the vehicle and configured to generate data; a packet network disposed within the vehicle and configured to transfer the data from the plurality of data sources to a central storage location within the vehicle; A processor installed in the vehicle, generating metadata corresponding to the data transferred from the plurality of data sources and stored in the central storage location; configured to transfer one or more selected portions of the sensor data to an external system external to the vehicle, the selected portions being selected based on the metadata corresponding to the data transferred from the multiple data sources and stored in the central storage location. Processor and An automobile data processing system comprising: [Item 41] Item 41. The vehicle data processing system of item 40, wherein the plurality of data sources includes two or more different data sources, each configured to produce a different type of data. [Item 42] Item 41. The automobile data processing system of item 40, wherein the data produced by the two or more data sources of the plurality of data sources are different types of data. [Item 43] 43. The vehicle data processing system of any one of items 40 to 42, wherein the processor is configured to select the one or more portions of the sensor data in response, at least in part, to an indication of one or more features of interest received from the external system. [Item 44] 43. The vehicle data processing system of any one of claims 40 to 42, wherein the processor is configured to generate the metadata by applying an artificial intelligence (AI) inference model to the data stored in the central storage location. [Item 45] 43. The automobile data processing system of any one of items 40 to 42, wherein, in generating the metadata, the processor is configured to tag one or more occurrences of one or more features of interest within the data. [Item 46] 43. The vehicle data processing system of any one of items 40 to 42, wherein in generating the metadata, the processor is configured to tag occurrences of the same feature of interest within two or more different types of the data or within data generated by two or more different data sources. [Item 47] generating data from a plurality of data sources distributed at different locations within a vehicle; transferring said data from said plurality of data sources to a central storage location within said vehicle via a packet network disposed within said vehicle; Using a processor installed in the vehicle, generating metadata corresponding to the data transferred from the plurality of data sources and stored in the central storage location; and transferring one or more selected portions of the sensor data to an external system external to the vehicle, the selected portions being selected based on the metadata corresponding to the data transferred from the multiple data sources and stored in the central storage location. and using the An automobile data processing method comprising: [Item 48] Item 48. The method of claim 47, wherein the plurality of data sources includes two or more different data sources, each configured to produce a different type of data. [Item 49] Item 48. The automobile data processing method according to item 47, wherein the data produced by the two or more data sources of the plurality of data sources are different types of data. [Item 50] 50. The method of processing vehicle data of any one of claims 47 to 49, wherein the step of selecting the one or more portions of the sensor data is performed at least in part in response to an indication of one or more features of interest received from the external system. [Item 51] 50. The method of claim 47, wherein generating the metadata comprises applying an artificial intelligence (AI) inference model to the data stored in the central storage location. [Item 52] 50. A method for processing automobile data according to any one of claims 47 to 49, wherein the step of generating the metadata comprises tagging one or more occurrences of one or more features of interest within the data. [Item 53] 50. A method for processing vehicle data according to any one of claims 47 to 49, wherein generating the metadata comprises tagging occurrences of the same feature of interest within two or more different types of the data or within data generated by two or more different data sources. [Item 54] an interface for communicating with a vehicle; Computer and Equipped with The computer, defining a model that identifies one or more specified features in data produced by one or more data sources for the vehicle; providing the model to a processor installed in the vehicle; receiving metadata from the processor in the vehicle, the metadata being generated by the processor using the model, the metadata tagging occurrences of the specified features in the data; Parsing the received metadata It is configured as follows: Automotive data analysis system. [Item 55] 55. The automobile data analysis system of claim 54, wherein the interface is configured to receive the metadata while the vehicle is connected to a docking station. [Item 56] 55. The automobile data analysis system of claim 54, wherein the interface is configured to receive the metadata from the vehicle regardless of whether the vehicle is connected to a docking station. [Item 57] 57. The automobile data analysis system of any one of claims 54 to 56, wherein the model includes an artificial intelligence (AI) inference model, and the computer is configured to train the AI ​​inference model and provide the trained AI inference model to the processor in the vehicle. [Item 58] 57. The automobile data analysis system of any one of claims 54 to 56, wherein the computer is configured to send a request for one or more selected portions of the stored data to the processor in the vehicle in response to the received metadata, receive the requested one or more selected portions from the processor, and analyze the one or more selected portions of the data. [Item 59] 57. The vehicle data analysis system of any one of items 54 to 56, wherein the computer is configured to request two or more selected portions of the stored data, the two or more selected portions being produced by two or more different data sources of the plurality of data sources. [Item 60] 57. The automotive data analysis system of any one of items 54 to 56, wherein the computer is configured to request two or more selected portions of the stored data, the two or more selected portions being associated with a same feature of interest. [Item 61] defining, in a computer external to the vehicle, a model that identifies one or more specified features in data produced by one or more data sources of the vehicle; providing the model to a processor installed in the vehicle; receiving, in the computer external to the vehicle, metadata from the processor in the vehicle, the metadata generated by the processor using the model, tagging occurrences of the specified features in the data; parsing the received metadata; The automobile data analysis method includes:

Claims

1. a storage subsystem disposed within the vehicle and configured to store at least data produced by one or more data sources of the vehicle; A processor installed in a vehicle, applying at least one model to the data stored in or en route to the storage subsystem that identifies one or more specified features of interest in the data to generate metadata that tag occurrences of the one or more specified features of interest in the stored data; exporting at least a portion of the metadata via a data link to an external system external to the vehicle, the metadata tagging the occurrences of the one or more specified features of interest prior to exporting the data including the one or more specified features of interest tagged with the metadata; after exporting the metadata to the external system and before exporting the data including the one or more specified features of interest to which the exported metadata relates, receiving a request from the external system, external to the vehicle, for one or more selected portions of the data determined by analysis by the external system based on the metadata, the selected portions of the data relating to particular features of interest tagged by the metadata generated at the vehicle using the at least one model and exported from the vehicle to the external system; a processor configured to provide, based on the request, from the processor installed in the vehicle to the external system a selected portion of the data encompassing the one or more specified features of interest, the selected portion of the data being less than all of the data produced by one or more data sources of the vehicle to reduce the amount of data transferred over the data link; An automobile data processing system comprising:

2. 2. The vehicle data processing system of claim 1, wherein the processor is configured to identify occurrences of the same feature of interest in the data produced by two or more different data sources of the plurality of data sources and, when generating the metadata, to tag the data to link together the occurrences of the same feature of interest that appear in the data produced by the two or more different data sources of the plurality of data sources.

3. 3. The vehicle data processing system of claim 2, wherein the data produced by the two or more data sources of the plurality of data sources is the same type of data.

4. 3. The vehicle data processing system of claim 2, wherein the data produced by the two or more data sources of the plurality of data sources are different types of data.

5. 3. The automobile data processing system of claim 2, wherein the two or more data sources of the plurality of data sources are located at different locations within the vehicle.

6. 5. The vehicle data processing system of claim 4, wherein the plurality of data sources includes at least two of a video camera, a speed sensor, an accelerometer, an audio sensor, an infrared sensor, a radar sensor, a lidar sensor, an ultrasonic sensor, or a distance measuring device.

7. 7. The vehicle data processing system of claim 1, wherein the processor is configured to provide two or more selected portions of the data to the external system, the two or more selected portions being produced by two or more different data sources of the plurality of data sources and associated with the same feature of interest.

8. 7. The motor vehicle data processing system of claim 1, wherein the processor is configured to select the portions of the metadata to export according to a selection criterion.

9. 7. The motor vehicle data processing system of claim 1, wherein the processor is configured to export the metadata without simultaneously exporting data corresponding to the metadata.

10. 7. The automobile data processing system of claim 1, wherein the storage subsystem is configured to store at least data produced by a plurality of independent data sources within the vehicle.

11. 7. The automobile data processing system of claim 1, wherein the storage subsystem is configured to store at least data produced by a plurality of data sources installed at different locations within the vehicle.

12. 7. The vehicle data processing system of claim 1, wherein the storage subsystem is configured to store two or more different types of the data.

13. 7. The motor vehicle data processing system of claim 1, wherein the processor is configured to receive the model from the external system.

14. the processor is configured to export the metadata while the vehicle is connected via a data link to a docking station, and then export the one or more selected portions in response to the request of the external system.

7. A vehicle data processing system according to any one of claims 1 to 6.

15. 7. The motor vehicle data processing system of claim 1, wherein the processor is configured to export the metadata regardless of whether the vehicle is connected to a docking station.

16. 7. The vehicle data processing system of claim 1, wherein the at least one model includes at least one pre-trained artificial intelligence (AI) inference model, and the processor is configured to generate the metadata by applying the at least one pre-trained AI inference model to the data.

17. 20. The automobile data processing system of claim 16, configured to update training of the at least one pre-trained AI inference model within the vehicle.

18. 7. The vehicle data processing system of claim 1, wherein said storage subsystem includes a centralized storage device configured to store said data produced by a plurality of said data sources.

19. 20. The vehicle data processing system of claim 18, wherein the centralized storage device is configured to store the data produced by two or more data sources of different types among the plurality of data sources.

20. The processor, applying a second model to the data, the second model being different from the at least one model, the second model identifying one or more second features of interest within the data; adding second occurrences of the one or more second features of interest to the metadata according to the second model; and and (iii) providing to the external system one or more of: (i) metadata regarding only the occurrence of the feature of interest; (ii) metadata regarding only the second occurrence of the one or more second features of interest; and (iii) metadata regarding both the occurrence and the second occurrence.

7. A vehicle data processing system according to any one of claims 1 to 6.

21. 7. The motor vehicle data processing system of claim 1, wherein the processor is configured to store the metadata in a structured form within the storage subsystem.

22. storing, in a storage subsystem located within the vehicle, at least data produced by one or more data sources of the vehicle; using a processor installed in the vehicle, applying at least one model to the data stored in or en route to the storage subsystem that identifies one or more specified features of interest in the data to generate metadata that tag occurrences of the one or more specified features of interest in the stored data; exporting at least a portion of the metadata via a data link to an external system external to the vehicle, the metadata tagging the occurrences of the one or more specified features of interest prior to exporting the data including the one or more specified features of interest tagged with the metadata; after exporting the metadata to the external system and before exporting the data encompassing the one or more specified features of interest to which the exported metadata relates, receiving, by the processor installed in the vehicle, from the external system external to the vehicle, a request for one or more selected portions of the data determined by analysis by the external system based on the metadata, the selected portions of the data relating to particular features of interest tagged with the metadata generated at the vehicle using the at least one model and exported from the vehicle to the external system; and providing, based on the request, from the processor installed in the vehicle to the external system a selected portion of the data encompassing the one or more specified features of interest, the selected portion of the data being less than all of the data produced by one or more data sources in the vehicle to reduce the amount of data transferred over the data link; and using the An automobile data processing method comprising:

23. 23. The method of claim 22, wherein applying the at least one model comprises identifying occurrences of the same feature of interest in the data produced by two or more different data sources of the plurality of data sources, and generating the metadata comprises tagging the data to link together the occurrences of the same feature of interest that appear in data produced by the two or more different data sources of the plurality of data sources.

24. 24. The method of claim 23, wherein identifying the occurrences of the same feature of interest comprises identifying the occurrences in data provided by two or more of the plurality of data sources, the two or more data sources being located at different locations within the vehicle.

25. 25. A method for processing vehicle data according to any one of claims 22 to 24, wherein providing the selected portions comprises providing two or more selected portions of the data to the external system, the two or more selected portions being produced by two or more different data sources of the plurality of data sources and associated with the same feature of interest.

26. 25. The method for processing motor vehicle data according to any one of claims 22 to 24, wherein exporting the metadata comprises exporting the metadata separately from data corresponding to the metadata, selecting one or more interesting parts of the data based on the metadata, and exporting the selected one or more interesting parts of the data when the vehicle is connected to a docking station.

27. The vehicle data processing system of claim 1 , wherein the at least one model comprises an AI inference model received from the external system.

Images