Edge-Cloud Data Processing for Machine Learning Model Updates
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Locally distributed devices in intelligent industrial control systems face limitations in computational resources, restricting machine learning capabilities, and require significant data transfer to cloud infrastructures for enhanced processing, which increases communication network load.
Innovation Solution
A data processing system that computes model parameters locally using limited resources and filters relevant data for transfer to an external device, such as a cloud, where enhanced processing can occur, minimizing data transmission and leveraging greater computational resources for improved model accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Extent of automation
If machine learning is performed on local devices, then machine learning capability is improved, but computational resources are insufficient
Solution Approach 1:
The system divides machine learning tasks into two segments: local device performs data filtering and preliminary processing, while cloud infrastructure performs computationally intensive model training and parameter optimization. This segmentation allows each component to operate within its resource constraints while achieving overall system intelligence.
Solution Approach 2:
The patent introduces an intermediary communication layer between local devices and cloud infrastructure. Local devices filter and prepare data locally, then transmit only essential features to the cloud. The cloud processes this data and returns optimized model parameters, creating a coordinated intermediary system that bridges resource limitations.
2Extent of automation
If data is transferred to cloud infrastructure for enhanced machine learning, then machine learning performance is improved, but communication network load increases
Solution Approach 1:
Local devices extract and filter only the most relevant features and data from sensor measurements before transmission to the cloud. This extraction process eliminates redundant information and reduces the quantity of data that needs to be transmitted over the communication network while preserving the essential information needed for machine learning.
Solution Approach 2:
Data filtering and preliminary processing are performed locally on edge devices before transmission to the cloud. This preliminary action prepares and optimizes the data structure in advance, ensuring that only processed and condensed information is transmitted, thereby reducing network bandwidth requirements.
3Measurement precision
If all sensor data is transferred to cloud for processing, then model accuracy is improved, but computational overhead on local devices increases
Solution Approach 1:
The system extracts only the essential features and critical sensor data needed for accurate modeling, discarding redundant information locally. This selective extraction maintains model accuracy by preserving meaningful patterns while significantly reducing the computational overhead required for data processing.
Solution Approach 2:
Local devices perform quality filtering and selective data preparation based on their specific operational context and requirements. Each local device tailors its data processing to its immediate needs, optimizing the balance between local computational effort and the quality of data transmitted to the cloud for final model refinement.
Data Source
AI summary
The present disclosure provides an enhanced computation of a data model for an intelligent data processing device. The data processing device may be a device having limited computational resources. Accordingly, a system model for processing the data is computed in the local device. Additionally, an enhanced model may be computed in a remote device like a cloud or a data center. For this purpose, the cloud or datacenter is provided with filtered data for computing an enhanced model. The cloud or datacenter may compute an enhanced model and forward the respective model to the local device if the enhanced model is better than the model locally generated.