JPWO2020097221A5

JPWO2020097221A5 -

Info

Publication number: JPWO2020097221A5
Application number: JP2021524162A
Authority: JP
Publication date: 2022-10-24

Claims

A system for managing vehicle data, comprising:
(i) data associated with one or more remote entities requesting one or more subsets of said vehicle data; and (ii) one or more generating said one or more subsets of said vehicle data. a data repository configured to store data related to an application of a data repository, said data repository being local to said vehicle for which said vehicle data is collected or generated;
a knowledge base configured to store a machine learning-based predictive model and user-defined rules for determining data transmission rules, comprising: (i) selected portions of said vehicle data to be transmitted; a remote entity of said one or more remote entities for a) transmitting said selected portion of said vehicle data; and (iii) receiving said selected portion of said vehicle data. , the knowledge base and
a transmission module configured to transmit portions of the vehicle data based on the data stored in the repository and the transmission rules.

2. The system of claim 1, wherein the repository, knowledge base and transmission module are onboard the vehicle.

3. The system of claim 1, wherein the one or more remote entities comprise cloud applications, data centers, third party servers, or another vehicle.

2. The system of claim 1, wherein the data repository stores data indicative of availability of the one or more subsets of the vehicle data, transmission timing delays, data types of the relevant subsets of data, or transmission protocols. .

2. The system of claim 1, wherein the machine learning based prediction model is stored in a model tree structure.

6. The system of claim 5, wherein the model tree structure represents relationships between machine learning based predictive models.

6. The system of claim 5, wherein nodes of the model tree structure represent machine learning-based predictive models, the nodes comprising at least one of model architecture, model parameters, training datasets, or test datasets.

2. The system of claim 1, wherein the machine learning-based predictive model is generated by a modeler located in a data center.

9. The system of claim 8, wherein the machine learning based predictive model is trained and tested using metadata and the vehicle data.

9. The system of claim 8, wherein the model generator is configured to generate a predictive model usable for the vehicle, the predictive model stored in the knowledge base.

3. The system of claim 1, wherein the vehicle is a connected vehicle, a connected automated vehicle, or a connected autonomous vehicle.

2. The system of claim 1, wherein the vehicle data is processed by a pipeline engine comprising one or more functional components, the pipeline engine generating metadata.

13. The system of claim 12, wherein at least one of said one or more functional components is selected from a set of functions via a user interface.

At least one of said one or more functional components is configured to create a scenario data object, said scenario data object usable to specify a scenario in which said metadata is used. 13. The system of claim 12.

15. The system of claim 14, wherein the metadata is usable for retrieving the subset of vehicle data from the remote entity for training the machine learning-based predictive model.