Cognitive Factory
|
CogMaSh aims to create a manufacturing system that achieves similar levels of flexibility,
robustness and improvement through experience as found in human machine shops.
In the production of prototypes, customized products and small or mid-size series, human workers
with their problem solving abilities, experience and cognitive capabilities are still the single way to provide the required flexibility, adaptability and reliability.
|
|
Research Topics
- Plan-Based Control
- Manufacturing
- RPL
- Cognition
- Plan Optimization
Application Domain
- Cognitive Factory
Project Details
The realization of comparable cognitive capabilities in production environments, therefore, is a step towards a Cognitive Factory that merges the advantages of automated and human operated production systems. In the Cognitive Factory, workflows are no longer rigid and hardwired. Instead, the system is equipped with its own online planning capability and may optimize its actions autonomously. Products and components in the Cognitive Factory perceive their own environment, store data and interact with other components of the production system. This will result in an improved quality assurance and true track & trace functionalities during production. Furthermore, the production system will improve its own production plans during runtime based on experience as well as by acquiring and continuously adapting internal system models. Process plans are then transformed online into control strategies, which realize the autonomous production for the momentary specifications. In the field of machining planning, a new feature-based grammar will be developed, that will help to enable the online generation of alternative process plans for a given part model and performance criteria. First test implementations have been made for an integrated Flexible Manufacturing System (FMS), which includes substantial parts and capabilities of an industrial production system. In the test bed, the autonomous production of a customized toy maze was demonstrated. The findings of the project will be applicable to a broad range of production systems in small and mid-size series.
