Product Assurance Runtime
The Product Assurance Runtime component focuses on ensuring product quality through data obtained in the manufacturing process and quality inspections. It consists of two tasks: Product Quality Prediction and Product Quality Supervision, both utilizing AI and Big Data techniques. This documentation specifically covers the Quality Prediction Designer, which allows the design and training of machine learning models for predicting product quality. The component integrates machine learning libraries and analytical tools to prepare and select trained models.
Benefits
Dataset management
Friendly user interface to train and optimize a model
Experiments tracking
Visualization tools to analyse the model
Save results from experiments
Export the model for run-time prediction
Product Quality Prediction Task
- Objective: Provide real-time predictions of product quality variables in a manufacturing process using Machine Learning Models and Big Data techniques.
- Function: Predicts product quality variables through the integration of a ML engine and the Quality Prediction Designer web UI.
Training Component
- Purpose: Build and train Machine Learning models for product quality prediction.
- Data Manager and Pre-processing
- Model Configuration and Run Experiment
- Analysis Tools
Inference Component
- Objective: Use trained models to make predictions in real-time scenarios.
- Process: Export the trained model as a model.zip archive, upload it using the Model Deployment Manager sub-component, and deploy it through the AI Analytics Runtime component.
Model Deployment Manager
- Function: Manage the deployment of trained models by handling the upload and integration process.
- Role: Allows users to export and deploy trained models for making predictions in real-time scenarios within the AI Analytics Runtime component.
Additional resources
Training Academy
Software Documentation
Read our easy to follow documentation to learn how to use the i4 Components.