Edge Intelligence Platform for Real-Time IoT Sensor Analytics
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Traditional cloud computing infrastructure is inadequate for handling the vast amounts of data generated by industrial machines due to latency, bandwidth, and cost issues, as well as the need for real-time decision-making and predictive maintenance in industrial IoT applications.
Innovation Solution
The implementation of an edge computing platform that enables intelligence at the source of data generation, using a software layer connected to a local-area network, with a repository of services and data processing engines, and a highly expressive analytics engine for real-time analytics and machine learning, allowing for efficient data processing and reduced bandwidth requirements.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of time
If all sensor data is sent to cloud storage, then centralized data processing is achieved, but latency increases and real-time decision making is compromised
Solution Approach 1:
The patent segments the centralized cloud computing architecture into distributed edge computing nodes deployed at multiple locations closer to data sources. This segmentation enables local data processing and analytics, reducing the time delay associated with transmitting all data to a centralized cloud data center while maintaining the benefits of distributed intelligence.
Solution Approach 2:
The patent introduces a new spatial dimension by deploying edge computing infrastructure at multiple geographic locations between the data sources and the central cloud. This dimensional change creates a hierarchical architecture where data can be processed locally at the edge before being aggregated centrally, thereby reducing latency for time-sensitive operations.
2Quantity of substance
If high bandwidth is provided for data transmission, then all sensor data can be transmitted, but cost becomes prohibitive
Solution Approach 1:
The patent extracts and processes data locally at edge computing nodes before transmission to the cloud. By performing filtering, aggregation, and preliminary analytics at the edge, only essential and processed data needs to be transmitted over the network, significantly reducing the volume of data transmission and associated bandwidth costs.
Solution Approach 2:
The patent implements partial data transmission by selectively sending only the most critical and processed data to the cloud, rather than transmitting all raw sensor data. This partial action approach reduces network bandwidth consumption while still providing sufficient data for cloud-based analytics and historical analysis.
3Reliability
If connectivity is ensured for cloud communication, then data can be transmitted reliably, but the system becomes less resilient to connection failures
Solution Approach 1:
The patent implements preliminary data processing and analytics at edge computing nodes before data needs to be transmitted to the cloud. By performing computations locally in advance, the system can maintain operational reliability even when connectivity is interrupted, as critical functions continue to operate using locally processed data.
Solution Approach 2:
The patent introduces edge computing nodes as intermediary systems between data sources and the central cloud. These intermediaries buffer and process data locally, decoupling the reliability of data transmission from the reliability of continuous cloud connectivity. The edge nodes serve as mediators that maintain system operation during connection failures.
4Productivity
If real-time analytics are performed at the edge, then immediate automated responses are enabled, but more computing resources are needed at distributed locations
Solution Approach 1:
The patent implements containerized applications that can be deployed across multiple edge devices with varying computational capabilities. These universal containers enable real-time analytics to be performed on devices with different resource profiles, allowing the system to leverage available computing power efficiently while maintaining the ability to perform immediate automated responses.
Solution Approach 2:
The patent dynamically adjusts analytics processing parameters and complexity based on the computational resources available at each edge device. By changing processing parameters such as analytics depth, data sampling rates, and model complexity, the system enables real-time operations on resource-constrained devices while maintaining operational efficiency.
Data Source
Figure 1~2
Figure 3
Figure 4
AI summary
A method for enabling intelligence at the edge. Features include: triggering by sensor data in a software layer hosted on either a gateway device or an embedded system. Software layer is connected to a local-area network. A repository of services, applications, and data processing engines is made accessible by the software layer. Matching the sensor data with semantic descriptions of occurrence of specific conditions through an expression language made available by the software layer. Automatic discovery of pattern events by continuously executing expressions. Intelligently composing services and applications across the gateway device and embedded systems across the network managed by the software layer for chaining applications and analytics expressions. Optimizing the layout of the applications and analytics based on resource availability. Monitoring the health of the software layer. Storing of raw sensor data or results of expressions in a local time-series database or cloud storage. Services and components can be containerized to ensure smooth running in any gateway environment.