Business Area
HOME Business Area AI Smart Factory Solution
AI Smart Factory Solution
This solution can be integrated and managed the entire process with only a few managers
Have you thought about what is needed most in the manufacturing process when introducing a smart factory?
Maybe it's a management.
Smooth and systematic management, such as production management, process management, and inventory management, can prevent production target disruptions and quality issues in advance.

To address these concerns, We propose TWIM's AI Smart Factory solution.
TWIM's AI smart factory solution can integrate and manage the entire process with only a few managers by accurately delivering and predicting the execution status of production plans in real time to help workers and managers make fast decisions and standardizing specifications such as product-specific inspection equipment, manipulation and monitoring.

Big Data

Overview
It analyzes and stores a large amount of structured and unstructured data (Legacy System) occurring in the manufacturing process in a big data warehouse in real-time, and manages the data so that it can be used directly in an artificial intelligence system.
Need more information or consulting? Contact us!
Sales Team in TWIM Corp.
Features
  • Establish the system to detect an abnormal signals (packets, system logs, etc.) using network traffic information
  • Real-time diagnosis and prediction of problem processes after immediate detection of changes in process capability index through real-time process status monitoring
  • Multi-dimensional analysis of process data improves production efficiency by pre-detecting quality problems and improving defect rates
Introduction Effects
  • Advances in monitoring and process control in manufacturing production processes: Production data visualization and micro-process control are possible for a long time by actively utilizing sensors and applying data collection/archival technology to support them.
  • Advances in analysis and feedback reflection of production processes: Improvements through analysis and feedback reflection of production processes can be made with data analysis reports and objective interpretation by engineers
  • Gain new power to improve productivity and efficiency: Big data and AI are the driving forces behind the Fourth Industrial Revolution, driving manufacturing productivity and efficiency
  • Tertiarization of manufacturing: It is possible to build a small-scale/multi-variety production system by utilizing all the information between customer demand and manufacturer's supply
Applicable Cases
Defective identification service through display process data analysis
  • Definition: Classification AI algorithm service that aims to determine the quality/defect of the product through facility data analysis that occurs in the display process.
  • Necessity: To reduce unnecessary costs for determining quality/defect in the manufacturing process.
  • Effects:
    - Reduce costs by minimizing the inspection process necessary for determining qualified/defective products by analyzing facility data
    - Increase productivity by enabling accurate and fast judgment with deep learning models
SMOTE (Synthetic Minority Over-sampling Technique)
  • · Solve imbalances by sampling multiple classes and synthesizing new decimal instances by storing existing decimal samples
  • · Overfitting can be prevented to some extent by creating new samples in combination with existing samples, not regenerating the same sample.
Ensemble
  • · By combining the prediction results of various models, predicted values with higher reliability than when analyzed with a single model are obtained.
Confusion Matrix
  • · The purpose of learning classification models is to classify the given data according to their intentions.
  • · Evaluate scales such as accuracy, precision, and reproducibility.
Raw Material Recommendation Services
  • Definition: This is a service that learns a product production key indicator prediction artificial intelligence model based on facility sensors, factory-run big data, and recommends raw materials by utilizing learned production key indicator prediction artificial intelligence models.
  • Necessity: Economic, stable and eco-friendly operation of production using artificial intelligence technology is required.
  • Effects:
    - Improve economic operation efficiency of production facilities, maintain stable operation, and induce eco-friendly operation.
    - Flexible propose the operation plan for production facilities
    - Avoid subjective experience and passive facility operation of existing managers and enable objective data and automatic facility operation.
Applications
  • Defective identification service through display process data analysis: Display process area
  • Solar power forecasting service through solar radiation: power plant or energy business area
  • Service for predicting yields through the growth environment: Agricultural area
  • Raw Material Recommendation Services: Any manufacturing area
Need more information or consulting? Contact us!
Sales Team in TWIM Corp.