Artificial Intelligence in Automation

The descibed topics are relate to a project work, bachelor or master thesis and can be started immediately. The following topics are currently available:

Industrial Communication

As a part of the vision of Industry 4.0 the interconnection of components will became more important because of the growing complexity of cyber-physical plants.
The research field of industrial communication provides elementary prerequisite for the vision of Industry 4.0. Special challenges are the communication in the industrial context and the necessary real-time capabilities, robustness and reliability in industrial applications.

More:  Research field Industrial Communication

  • DEVEKOS: Theses are offered in the following areas: Modeling and code- synthesis. Project description

Engineering und Configuration

Today’s fast changing markets require fast changing products, and adaptable production plants, the so-called Plug- and-Produce (PnP) paradigm. Nowadays, the automation systems become more and more the bottleneck for PnP, because each reconfiguration causes a high engineering effort for adapting the automation system.

Solutions are modular, self-organizing software structures in combination with intelligent assistant systems. Such approaches and the corresponding engineering tools are developed in several projects to support the user in the configuration and planning task. For this, PnP solutions try to formalize human configuration and automation knowledge, i.e. by means of rules, semantics and ontologies. In each case, the assistant system develops an optimal strategy to adapt the automation solution, i.e. a strategy that minimizes manual engineering steps.

More: Research field Engineering and Configuration

Analysis & Diagnosis

Intelligent Analysis and Diagnosis are major components in the application areas Cyber-Physical Production Systems and “Industrie 4.0”. To detect anomalies and errors during operation or optimise the process, models of the plants are required. The effort to manually create these models is very high and not practicable in today’s modern, fast changing, complex systems.

To avoid the manual modelling effort, models are learned from the process data of the production systems and are then used for anomaly detection and root cause analysis.

The results of these procedures are presented to the operator of the production system through novel decision support systems. A decision support system hides the plant’s complexity from the operators and supports them to monitor the plant by, for example, providing possible error causes when anomalies are detected.

More: Research Analysis & Diagnosis

Running projects:


Natalia Moriz (Link) and the contacts of the projects.