Digital Twin Comparison for Edge O&M
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current digital twin technologies face limitations such as dependence on cloud computing, insufficient bandwidth and latency, reliance on OEM data, high computational intensity, difficulty in calibrating machine learning models, inability to prescribe corrective actions, vulnerability to the curse of dimensionality, and lack of independence from OEMs, which hinder their effectiveness in operations and maintenance applications.
Innovation Solution
An operations and maintenance system that employs a database subsystem for storing distinct digital twins, a sensor subsystem for collecting operational data, and a digital twin comparison subsystem to make maintenance decisions based on the comparison of outputs from these twins, incorporating artificial intelligence for weighted comparisons and ensemble modeling to reduce dimensionality and enhance transparency.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Power
If cloud computing is used for digital twins, then computational capability is enhanced, but dependence on cloud infrastructure and network bandwidth increases
Solution Approach 1:
The patent creates local copies of digital twin data and models on edge devices or local systems, enabling computational operations without continuous cloud dependency. This allows the system to maintain computational capability while reducing infrastructure dependence.
Solution Approach 2:
The system divides computational tasks between local edge devices and cloud infrastructure, segmenting functionality to reduce dependence on any single infrastructure while maintaining overall computational power.
2Measurement precision
If machine learning models are used for predictive maintenance, then prediction accuracy is improved, but computational intensity and calibration difficulty increase
Solution Approach 1:
The system dynamically adjusts model parameters and computational complexity based on data availability, operational context, and prediction requirements, reducing unnecessary computational intensity while maintaining accuracy.
Solution Approach 2:
The patent applies machine learning models selectively only when and where needed, rather than continuously, reducing overall computational intensity while maintaining prediction accuracy for critical maintenance tasks.
3Adaptability or versatility
If multiple digital twins are maintained for diverse objects, then system versatility is improved, but complexity and diversity management difficulty increase
Solution Approach 1:
The system implements a universal digital twin platform that can handle multiple object types through standardized frameworks, reducing the complexity of managing diverse twins while maintaining system versatility.
Solution Approach 2:
The patent applies tailored characteristics and parameters specific to each object type while using a common underlying architecture, allowing diverse digital twins to be managed through localized customizations rather than complete system complexity.
Data Source
AI summary
Operations and maintenance (O&M) system, and related methods, for a plurality of objects employing distinct digital twins. The O&M system comprises: a database subsystem for storing first and second distinct digital twins for each of the plurality of objects, each of the distinct digital twins having an identifier that associates it with one of the plurality of objects and which defines a virtual representation thereof. The system further includes a sensor subsystem operative to obtain operational data for the plurality of objects, and a digital twin comparison subsystem operative to compare outputs of the at least first and second distinct digital twins for each of the plurality of objects; the output of each distinct digital twin is a function of the operational data for its associated object, and the O&M system makes an operational or maintenance decision with respect to an object as a function of the comparison.