Edge-Cloud Data Analysis for Adaptive Industrial Applications
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Industrial plants face challenges in quickly generating customized data analysis applications to improve production efficiency due to limitations in network transmission and the reliance on domain experts for specific algorithms, and existing solutions either rely on cloud-based mechanisms that are inefficient or local edge devices that lack sufficient computing capacity.
Innovation Solution
A data analysis method and system that performs analysis on edge devices, collecting key performance indicators, evaluating application performance, generating upgrade requirements, and sending them to the industrial cloud for context data collection and application updates, thereby leveraging the computing power of both edge devices and industrial clouds.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If all data is sent to cloud terminal for data analysis, then data analysis accuracy is improved, but network transmission burden increases and response time deteriorates
Solution Approach 1:
The patent segments the data analysis system into edge devices and cloud terminals. Edge devices perform preliminary data collection, filtering, and local analysis, while only essential data is transmitted to the cloud for comprehensive analysis. This segmentation reduces network transmission burden and enables faster local responses while maintaining overall analysis accuracy.
Solution Approach 2:
The patent implements preliminary data processing and filtering at the edge device before transmission to the cloud. By performing initial data collection, validation, and preprocessing locally, the system reduces the volume of data requiring network transmission and cloud processing, thereby improving response time while preserving data analysis quality.
2Productivity
If specific algorithms are used for data analysis, then data analysis effectiveness is improved, but dependency on domain experts increases and system flexibility deteriorates
Solution Approach 1:
The patent implements a multi-functional data analysis system that combines domain-specific algorithms with general-purpose machine learning models. The system can adaptively select and combine different analysis methods based on the specific industrial scenario, reducing dependency on fixed domain expert knowledge while maintaining analysis effectiveness and improving system flexibility.
Solution Approach 2:
The patent employs parameter-adjustable algorithms that can be dynamically configured based on industrial data characteristics. By allowing parameter optimization through data-driven approaches rather than fixed domain expert settings, the system maintains high analysis effectiveness while gaining flexibility to adapt to different industrial scenarios without requiring constant expert intervention.
3Loss of time
If data analysis is performed on edge devices, then response time is improved, but computing capacity limitation increases analysis difficulty
Solution Approach 1:
The patent implements partial data analysis at the edge device, focusing on time-critical functions such as real-time monitoring, anomaly detection, and immediate control responses. By performing only the essential analysis locally and transmitting detailed data to the cloud for comprehensive processing, the system achieves fast response times for critical functions while avoiding the complexity of implementing full analysis capabilities on resource-constrained edge devices.
Data Source
AI summary
A data analysis method, device and system are disclosed. In an embodiment, the method includes performing data analysis on the devices by an application by collecting at least one key performance indicator of the devices, estimating the performance of the application according to the result of the data analysis, generating an application upgrade requirement on the basis of the performance estimation result, and sending the upgrade requirement to an industrial cloud; receiving a context data requirement generated by the industrial cloud based upon of the upgrade requirement, collecting context data from the devices according to the context data requirement, and returning the context data to the industrial cloud; and downloading from the industrial cloud the application updated on the basis of the context data. The data analysis mechanism reduces the development process of custom and special applications by automatically upgrading to adapt to changes in field application conditions.