kubwa in Energy Industry and Data Analysis
• Resolving innovative energy problems using AI technology following the explosive increase in global energy consumption
• Provide efficient energy-saving solutions AI-based for facility operation that utilizes energy data
• Energy efficiency by applying AI and big data technologies in the value chain stage of energy production, supply and consumption
-
2014
Expected to develop ICT utilization and convergence technology with the energy sector
-
2017
New energy-related data business expansion of IoT, AI public data opening start, automobile comprehensive information, national energy
-
2019
Digitalization of the fast-growing energy industry, and increased utilization of data analysis in energy production, consumption, utilization, and storage
-
2020
kubwa, orgnized data analysis business energy sector
-
2021~
Dvelope eco-friendly AI that meets increased energy efficiency through energy data analysis and the Korean Green New Deal policy
Application Case 1. Resource Management - Leak Detection
• Water pipe leak detection algorithm through AI model
• Minimize damage through proactive measures by predicting leaks in advance
• Efficiently measure leak location, leak status, and leak volume
-
1
Design
AI-based intelligent leak integration
Solution platform design,
AI-based pipe acceleration signal abnormal detection development,
Leak location prediction development -
2
Development
AI-based leak data prediction platform,
AI-based leak location calculation,
Calculation of leaks and leaks -
3
Validation
AI-based leak platform and model field validation and optimization
-
4
Commercialization
Commercialization of AI-based leak detection platform and algorithm
Application Case 2. Energy Management – Motor Machine Diagnosis
• Build AI motor diagnosis predictive maintenance system
• Real-time diagnosis by predicting motor failure with an AI model through vibration sensor data
Data Exploration
Sensor data collection
vibration frequency
Data Cleansing
Vibration & current data
Check data distribution
Raw data selection
Normal/abnormal frequency confirmation
Labeling data
Feature Extraction
Feature engineering for time-based and spectrum-based
AI Model & Prediction
Selection of algorithms such as XGBoost, LigthGBM, SVM, Neural Networks, etc.
AI learning and fault diagnosis
AI Deployment
Building motor diagnosis dashboard