Service Management Method, Platform, Service Delivery System, and Computer Storage Medium

Patent metadata

Field	Value
Publication number	US-2024168825-A1
Application number	US-202318282802-A
Country	US
Kind code	A1
Filing date	Jan 9, 2023
Priority date	Jan 25, 2022
Publication date	May 23, 2024
Grant date	—

Abstract

A service management method, a platform, a service delivery system, and a computer storage medium. The service management method includes: using a service delivery platform to obtain edge node monitoring information, and selecting an available edge node from the obtained monitoring information; creating one or a plurality of transactions, each of the transactions including one or a plurality of artificial intelligence (AI) services; using the service delivery platform to issue the created transaction(s) to the available edge node. The embodiments of the present disclosure realize a cloud-native edge node management method, the transaction issuing mode is efficient and convenient, the transaction management mechanism is efficient and flexible, and the use of AI service orchestration ensures flexible construction, convenience and controllability of transactions.

First claim

1 . A service management method, comprising:
acquiring monitoring information of an edge node by a service delivery platform, and selecting an available edge node from the obtained monitoring information; creating one or more transactions, wherein each created transaction comprises one or more artificial intelligence (AI) services; and distributing the created transactions to the available edge node through the service delivery platform.

2 . The service management method of claim 1 , wherein each created transaction further comprises at least one of: an AI service orchestration script, an image acquisition device corresponding to the transaction, or the available edge node corresponding to the transaction.

3 . The service management method of claim 1 , further comprising: initializing the available edge node.

4 . The service management method of claim 3 , wherein initializing the available edge node comprises:
selecting an application that is a control management program for defining an operation of the edge node; acquiring a mirror list from the service delivery platform, and selecting a basic container mirror from the mirror list; and according to the selected application and the basic container mirror, constructing an application container mirror by the service delivery platform and storing the application container mirror in a container warehouse, and controlling, by the service delivery platform, the available edge node to download the application container mirror, and construct and start the application container.

5 . The service management method of claim 4 , wherein the application comprises receiving the distributed transactions, dynamically loading transactions, parsing service orchestration transactions and starting thread to run transactions.

6 . The service management method of claim 1 , wherein creating the transaction comprises:
selecting one or more AI services; generating an AI service orchestration script according to the selected one or more AI services; selecting one or more image acquisition devices; and creating the transaction using the one or more AI services, an AI service orchestration script, the selected image acquisition devices and information of the available edge node.

7 . The service management method of claim 1 , further comprising:
acquiring a transaction distribution or update status of the available edge node by the service delivery platform.

8 . The service management method of claim 1 , further comprising:
controlling applications on the edge node by the service delivery platform, wherein the controlling comprises at least one of: creation, start, stop or update.

9 . The service management method of claim 1 , further comprising:
establishing an association between an application and the transaction of the edge node, wherein the application is a control management program for defining an operation of the edge node; and arranging, on a plurality of the edge nodes, a unified shared storage configured to store configuration data of the application and transaction of the plurality of the edge nodes.

10 . The service management method of claim 9 , wherein based on a determination that the available edge node fails, the method further comprises:
detecting whether a first edge node as an idle edge node exists, and migrating the application and transaction of the failed edge node to the first edge node when the first edge node exists.

11 . The service management method of claim 1 , further comprising:
acquiring and displaying an AI service processing result of one or more available edge nodes.

12 . The service management method of claim 11 , wherein
acquiring the AI service processing result of one or more available edge nodes is through any one or more of the following service channels: Message Queuing Telemetry Transport, Remote Dictionary Server, distributed publish and subscribe messaging systems or WebSocket.

13 . The service management method of claim 1 , wherein one of the edge nodes is configured to run a plurality of transactions in parallel, and each of the transactions is initiated as a thread.

14 . A service management platform, comprising a memory and a processor coupled to the memory, wherein the processor is configured to perform acts of the service management method of claim 1 based on instructions stored in the memory.

15 . A service delivery system, comprising the service management platform of claim 14 and further comprising a service delivery platform and one or more edge nodes, wherein:
the service management platform, the service delivery platform and the edge nodes are interconnected via a network;
the service delivery platform is configured to monitor and control one or more edge nodes, and distribute transactions created by the service management platform to one or more edge nodes; and
the edge node is configured to receive the transactions distributed by the service delivery platform, process the transactions, and return a processing result to the service management platform.

16 . The service delivery system of claim 15 , further comprising one or more image acquisition devices, wherein:
the edge nodes and the image acquisition devices are interconnected via the network; and the edge node processing the transactions, comprises: according to a received transaction, acquiring an image or video data collected by a corresponding one or more image acquisition devices, and processing the image or the video data.

17 . The service delivery system of claim 15 , wherein:
the edge node is configured to run a plurality of transactions in parallel, and the transactions comprises customer face registration transactions, customer identification transactions, shop arrival frequency analysis transactions, and stranger reminder transactions; the customer face registration transactions comprise a face detection service, a face scaling service, a face key point detection service, a face alignment service, a face feature extraction service and a face feature storage service; the customer identification transactions comprise a video decoding service, an image scaling service, a target detection service, a target tracking service, a face detection service, a face scaling service, a face quality evaluation service, a face alignment service, a face attribute determination service, a face feature extraction service, a face retrieval service and a stranger registration service; the shop arrival frequency analysis transactions comprise a record uploading service and a record statistics service; and the stranger reminder transactions comprise a stranger arrival statistics service and a reminder service.

18 . A non-transitory computer readable storage medium, stored thereon a computer program, wherein when the computer program is executed by a processor, the service management method of claim 1 is implemented.

19 . The service management method of claim 10 , wherein based on a determination that the available edge node fails, the method further comprises:
based on a determination that there is no first edge node, detecting whether there is a second edge node, an application version of the second edge node is consistent with an application version of the failed edge node and the second edge node has surplus resources for receiving the transaction of the failed edge node, deleting the association between the application and the transaction of the failed edge node when the second edge node exists, and migrating the transaction of the failed edge node

Assignees

• Boe Technology Group Co Ltd

Inventors

• Wang Xiaohong

Classifications

• G06N3/08

  Learning methods · CPC title

• G06N20/00

  Machine learning · CPC title

• G06F11/3006

  where the computing system is distributed, e.g. networked systems, clusters, multiprocessor systems (multiprogramming arrangements G06F9/46; allocation of resources G06F9/50) · CPC title

• G06F9/4856

  resumption being on a different machine, e.g. task migration, virtual machine migration (G06F9/5088 takes precedence) · CPC title

• G06F9/5077

  Logical partitioning of resources; Management or configuration of virtualized resources (specific details on emulation or internal functioning of virtual machines G06F9/455) · CPC title