Edge-Cloud AI Collaboration System for Dynamic Data Processing
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing edge-cloud collaboration systems face inefficiencies in terms of cost and resource utilization due to the need for high-specification edge devices for data processing, and privacy concerns arise from transmitting all edge data to the cloud server, which may include personal data.
Innovation Solution
A collaboration system between edge devices and cloud servers that includes a service coordinator, edge inferencer, and cloud inferencer, managed by a collaboration manager to optimize data flow and configuration based on log information and resource availability, allowing edge data to be processed at appropriate locations depending on type and purpose, with dynamic updating of artificial intelligence models.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If all edge data is transmitted to the cloud server for processing, then data analysis capability is improved, but privacy concerns increase and communication traffic increases
Solution Approach 1:
The system segments data processing into two parts: edge devices perform preliminary processing and filtering of edge data, while only necessary data is transmitted to the cloud server. This segmentation reduces the amount of data transmitted to the cloud, thereby reducing privacy risks and communication traffic while maintaining analysis capability.
Solution Approach 2:
The patent applies local quality by enabling edge devices to perform data processing locally rather than transmitting all data to the cloud. Each edge device processes its own data according to local needs and policies, which reduces privacy exposure and communication overhead while maintaining processing capability at the edge.
2Measurement precision
If all edge data is transmitted to the cloud server for processing, then data analysis capability is improved, but communication traffic increases
Solution Approach 1:
The system extracts and processes data locally at edge devices before transmission to the cloud. By taking out the preliminary processing step from the cloud and placing it at the edge, the system reduces the volume of data that needs to be transmitted, thereby reducing communication traffic while maintaining analysis capability.
Solution Approach 2:
The patent implements preliminary action by performing data processing and filtering at edge devices before transmission to the cloud server. This preliminary processing reduces the amount of data that needs to be transmitted, thereby reducing communication traffic and costs while maintaining the ability to analyze data in the cloud.
3Object-affected harmful factors
If edge computing is used to process data locally, then privacy is protected and communication traffic is reduced, but device specification requirements increase
Solution Approach 1:
The system implements universality by providing a cloud server that offers multiple processing functions and models that can be dynamically deployed on edge devices. This allows edge devices with varying specifications to access sophisticated processing capabilities through the cloud, reducing the need for high-specification hardware while maintaining privacy protection and processing capability.
Solution Approach 2:
The patent introduces a cloud server as an intermediary that provides processing capabilities to edge devices. Instead of requiring edge devices to have high specifications for local processing, the cloud server acts as a mediator that offers sophisticated analysis functions, thereby reducing device specification requirements while maintaining processing capability.
4Adaptability or versatility
If artificial intelligence models are applied to edge devices and cloud servers, then data analysis versatility is improved, but model configuration and updating complexity increases
Solution Approach 1:
The system implements dynamics by enabling dynamic configuration and updating of AI models on both edge devices and cloud servers. The collaboration manager can dynamically adjust model parameters, select different models, and update configurations based on changing requirements, thereby maintaining versatility while simplifying the complexity through centralized management.
Solution Approach 2:
The patent introduces feedback mechanisms where the collaboration manager monitors performance and usage data from edge devices and cloud servers, and uses this feedback to automatically configure and update AI models. This feedback loop simplifies model management by enabling automated adjustments based on real-world performance, reducing manual configuration complexity while maintaining versatility.
Data Source
AI summary
The present disclosure may provide a system and a method for collaboration between an edge and a cloud, which enable edge data to be inferred at a proper location among an edge device and a cloud server according to the type and the purpose of the edge data, and enable configurations of the edge device and the cloud server to be easily updated, wherein the edge device comprises: a service coordinator which is in charge of a flow in a collaboration service of the edge data collected by at least one sensor for each of multiple services provided by the collaboration system; and an edge inferencer which may include at least one edge artificial intelligence model for receiving an input of the edge data as input data for inference and outputting an inference result for each of at least one service among the multiple services, and the cloud server comprises: a cloud inferencer including at least one cloud artificial intelligence model for receiving an input of the edge data as input data for inference and outputting an inference result for each of at least one among the multiple services; and a collaboration manager for changing a configuration of at least one of the edge inferencer and the cloud inferencer on the basis of log information of the edge inferencer and log information of the cloud inferencer.