Asset Management Workflow With Domain-Specific Sensor Mapping
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing predictive maintenance solutions for industrial assets lack customization for specific domains, resulting in low accuracy and reliability, and fail to automatically determine required sensors and integrate new assets seamlessly, leading to complex and time-consuming workflows.
Innovation Solution
A system and method for managing assets in a technical installation that involves receiving requirements, generating user profiles, selecting relevant assets, mapping assets to sensing units, extracting domain knowledge, determining performance indicators, and automatically training machine learning models to enhance predictive maintenance.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If a generic predictive analytics solution is applied to every requirement, then the solution can be implemented quickly and with low cost, but the accuracy and reliability of asset management deteriorate due to lack of customization for specific domains
Solution Approach 1:
The system segments the asset management solution into domain-specific modules (continuous process industry, discrete industry, building space, manufacturing factory) each with customized workflows, sensors, and machine learning models tailored to specific industrial domains, thereby maintaining high accuracy while enabling selective implementation
Solution Approach 2:
The system allows dynamic configuration of parameters including sensor selection, workflow definitions, and model hyperparameters based on domain-specific requirements, enabling customization without requiring complete solution redesign for each domain
2Reliability
If manual configuration is performed for each individual asset including sensor selection and workflow setup, then the solution can be customized to specific requirements, but the complexity and time required for implementation increases significantly
Solution Approach 1:
The system pre-configures domain-specific workflows, sensor mappings, and machine learning models for different industrial sectors beforehand, so that when a new asset is added, the system can automatically apply relevant pre-configured settings rather than requiring manual configuration from scratch
Solution Approach 2:
The system automatically determines required sensors, selects appropriate workflows, and configures models based on asset type and domain knowledge, enabling self-service configuration that reduces manual intervention while maintaining customization
3Measurement precision
If domain knowledge and sensor data are used to train machine learning models for each asset, then the predictive accuracy improves, but the time and resources required for model training and retraining increase
Solution Approach 1:
The system merges training efforts across multiple assets by identifying common patterns and transferring learned knowledge between similar assets within the same domain, reducing redundant training while maintaining asset-specific predictive accuracy through domain-adaptive models
4Reliability
If new assets are integrated into the system manually, then the configuration can be optimized for each asset, but the integration process becomes complex and time-consuming
Solution Approach 1:
The system automatically integrates new assets by detecting asset type, selecting appropriate domain-specific workflows and sensors, and configuring models based on asset characteristics, enabling seamless self-service integration without manual intervention while maintaining configuration accuracy
Data Source
AI summary
A system, apparatus and method for managing plurality of assets in technical installation is provided. The method includes receiving, by a processing unit, a set of requirements for managing the plurality of assets, selecting one or more assets from the assets based on the received set of requirements, mapping the one or more assets to corresponding sensing units, extracting information associated with the selected one or more assets and the received set of requirements, determining performance indicator based on information extracted from the knowledge base and mapped sensing units, defining workflow to be executed based on the determined at least one performance indicator, selecting configured machine learning model from a set of machine learning models based on the defined workflow, and determining an outcome of the selected machine learning model based on the received set of requirements.