Master / Bachelor thesis: Generate a knowledge base from the digital twin
Expert systems are programs, that can replicate expert knowledge and inference capabilities of qualified experts in specific application areas. They consist of a knowledge base and an inference engine. Compared to AI methods, expert systems do not rely on the availability of a large amount of data and the reasoning behind the algorithms choice is evident.
Different implementations for the knowledge base exist. Among them are Petri nets, IF THEN rules or case studies and corresponding action recommendations.
In a previous thesis a Kinect was used to mirror the state of a Fischertechnik robot onto the digital twin. Also, sensor data can be gathered. The student shall “teach” the digital twin actions for various states (i.e. if the user approaches the robot it should stop). Relevant data for states and actions are stored and transformed to an ontology / rule. To achieve this, the student needs to choose an appropriate implementation of the knowledge base. The knowledge base should be extendable to add further state-action rules in the future. If a similar state is detected in the future, an expert system should choose the appropriate action.
Programming knowledge required.
Keywords: expert systems, digital twin
Betreuer: Hüsener
