Cognitive Systems Engineering

The Cognitive Systems Engineering research group led by Dr.-Ing. Sebastian Wrede was established in January 2009 at the Cluster of Excellence in Cognitive Interaction Technology CITEC (EXC 277) and Bielefeld University’s Faculty of Technology.

We investigate software-oriented systems engineering principles to support the development and operation of interactive cognitive systems. Tasks in these application domains are inherently complex and raise interesting research questions. For instance, in many existing systems, the behaviour among interacting system components is only imprecisely or implicitly defined at design and at run-time, which complicates the system integration process and makes a meaningful verification impossible. A better description, understanding and more efficient development process at the system- and component-level is one of the primary ambitions of the cognitive systems engineering group. We explore the application of model-driven systems engineering methodology using general purpose (i.e. BPMN, SysML, UML, …) and composable domain-specific modelling languages based on state-of-the-art language workbenches. Here, we focus on executable models that are either interpreted or transformed to source code tailored to a specific target environment. Furthermore, we employ dynamic simulation and data-driven machine learning techniques to verify system behaviour at design and at run-time. Ultimately, we search for methods that efficiently combine model- and data-driven approaches to support the complete software lifecycle of cognitive systems.

We try to bring the development of software engineering methods close to actual application in order to verify them with first hand experiences. For this purpose, we provide the required tools (often as open-soure projects) and services and maintain these. The methods and tools are explored in different collaborative research projects focusing on the development of cognitive systems with applications ranging from smart manufacturing to assistance systems in home environments or public spaces using different types of robots and interaction technology. Beyond the development and evaluation of our methods in fundamental research projects, we also participate in transfer-oriented industry projects through our association with the CoR-Lab of Bielefeld University. Exemplary projects are conducted in the leading-edge cluster it’s OWL on Industry 4.0 where we work on easily configurable robot systems and assistance systems for smart manufacturing jointly with our industry partners.