Streaming Analytics Source Operator Optimization
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current streaming analytics systems face challenges in optimizing and simulating real-time data streams from virtual models, particularly in dynamically adjusting to varying workloads and resource usage, without requiring exhaustive tracking of all physical asset changes.
Innovation Solution
The method involves using digital twin models as source operators to simulate and optimize data streams by dynamically changing the number and parameters of source operators, allowing for the refinement of streaming analytics applications and creation of new virtual models based on simulated real-time data from pre-existing models, without needing to track all changes in the physical assets.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If digital twin models are used as source operators to simulate data streams, then the ability to optimize and test streaming analytics applications is improved, but the complexity of setting up and managing virtual models increases
Solution Approach 1:
The patent creates virtual model instances that replicate physical asset behavior and data stream characteristics. These copies serve as source operators that generate simulated data streams matching the statistical properties and patterns of real assets, enabling testing and optimization without direct connection to physical systems.
Solution Approach 2:
The system dynamically adjusts parameters of virtual models including data generation rates, stream formats, and statistical properties to match different physical asset configurations. This allows the same virtual model framework to adapt to various asset types and operational conditions through parameter modification rather than structural changes.
2Measurement precision
If simulated real-time data from virtual models is used for refining streaming analytics applications, then the accuracy and performance of the analytics system is improved, but the time and resources required to set up simulation environments increase
Solution Approach 1:
Virtual model instances are pre-configured with asset metadata, data generation logic, and stream characteristics before actual analytics deployment. This preliminary setup creates ready-to-use simulation environments that can be quickly activated for testing and refinement without requiring extensive on-site configuration or physical asset connection.
Solution Approach 2:
The virtual model instances serve as intermediary components between physical assets and streaming analytics applications. They translate physical asset characteristics into simulated data streams that preserve statistical properties and patterns, enabling accurate analytics testing without direct physical asset involvement.
3Adaptability or versatility
If the number and parameters of source operators are dynamically changed to simulate varying workloads, then the realism and usefulness of the simulation is improved, but the computational overhead and system complexity increases
Solution Approach 1:
The system dynamically modifies the number of active virtual model instances and their operational parameters based on simulated workload conditions. Source operators can be added, removed, or reconfigured to represent different operational scenarios such as varying asset counts, data generation rates, and stream volumes, allowing realistic workload simulation without static configuration.
Data Source
AI summary
A computer system where pre-existing virtual models of physical assets, processes and/or computer system supply input data streams to a streaming analytics application through respective stream operators. The streaming analytics application uses this input data to make improvements to the code and/or configuration of the streaming analytics application itself and/or to create newly-created virtual model(s).