Symbolic methods and optimization

From training in the model to effective implementation.

Symbolic AI

What do we mean when we address research projects, projects and concrete solutions using symbolic AI?

Symbolic methods process symbols, e.g. in the form of logics, which are understandable and interpretable by humans. In contrast, subsymbolic AI is distinct from this: Here, we are dealing with many, unordered data points that cannot be easily interpreted. 

Example: Everyone can imagine something under the system state "good" (symbolic). From a time series of binary data it is very difficult for humans to find out the system state (sub-symbolic).

Guiding themes Symbolic procedures

The research activities and services of the group "Symbolic Methods and Optimization" are structured according to the four guiding themes:

  • Expert systems (Knowledge Modeling)
  • Diagnosis (Application field industry and SME, possible with small amount of production data)
  • Reconfiguration (Industry application field, redundancies necessary)
  • Explainable AI (Example project: KOSMOX)

Key topics of the Optimization division

The competencies along the guiding themes address three areas in application:

  • Optimization of production planning and control (Example project: SUPPORT)
  • Logistics optimization (Example project: IMAGINE)
  • Parameters optimization (Example project: KI4LSA)


Our range of services - from strategy to implementation:

  • Consulting
  • Scientific monitoring
  • Implementation

More information

Research topics and projects