INTELLIGENCE
Predictive models based on AI integrated into the Digital Twin
INTELLIGENCE Service Application__
The INTELLIGENCE application offers a platform for the deployment of artificial intelligence models integrated into the equipment’s Digital Twin, using machine data to solve specific problems for each use case.
Following simplified commands, users can compile a complete database, build the predictive model, train it with a specific objective, and test its performance directly on the equipment’s Digital Twin.
INTELLIGENCE includes various features to facilitate the implementation and management of AI models, as well as to
Repository of internally trained or imported models
Deployment of models on Edge & Cloud.
Anomaly detection and predictive maintenance.
LLM language model for knowledge extraction
FUNCTIONALITIES_
_ to cover a range of after-sales features
DATA COLLECTION
It implements the collection of machine data, with the frequency and quality required for the implementation of AI models.
The resulting datasets are securely identified and stored on the servers for future use.
AI MODEL REPOSITORY & DEPLOYMENT
It offers a repository of predefined AI models for various prediction and optimization use cases, which are made available to the user for specific implementation. These models can be deployed either in the cloud or at the Edge directly from the repository.
The models in the repository can serve different purposes, such as diagnostics and anomaly detection, predictive maintenance, reinforcement learning (RL) models, or LLMs (Large Language Models) for extracting knowledge from operational or status data.
AI MODEL CONFIGURATION & TRAINING
It offers a guided step-by-step configurator for defining the structure of the AI model and for training it. Users can start training a new model or import a previously trained one.
Training can be carried out using data collected directly from the machine (Edge data collection) or imported from other devices such as external loggers or datasets from other databases.
