Model training pipeline for finite state machines
- Project
- 19027 AIToC
- Type
- New service
- Description
The model training pipeline for finite state machines is part of AIToC's PAIPline platform for creating behavior models from real-world generated data. The model training pipeline automatically generats a behavior model in form of a finite state machine based on binary time series data. The binary data has to represent the states of production resource components.
- Contact
- Ireneus Wior, TWT GmbH Science & Innovation
- ireneus.wior@twt-gmbh.de
- Research area(s)
- Realistic and standardized production resource models
- Technical features
Modules with the following functionalities are included:
- Finite state machine generator: Generates a behavior model of the production resource component in form of a finite state machine.
- FMU exporter (in progress): Wraps a finite state machine model as a FMU container and exports it.
- Integration constraints
Needs a running instance of the PAIPline platform.
- Targeted customer(s)
Automotive industry, Manufacturing industry
- Conditions for reuse
On request.
- Confidentiality
- Public
- Publication date
- 29-02-2024
- Involved partners
- TWT GmbH Science & Innovation (DEU)