Aarhus University Seal / Aarhus Universitets segl

Research

In the Distributed Computational Intelligence (DCI) group, we perform research in both areas, computational intelligence and distributed systems. We consider the individual systems to have some sort of intelligence enabling them to operate autonomously and learn about themselves and their environment.

Self-awareness and networked self-awareness

With the advances of artificial and computational intelligence, we get devices and systems able to improve their performance towards predefined goals within specified boundaries. Computational self-awareness investigates psychological processes to be used in computing systems in order to enable them to learn about themselves and their environment. Since systems usually not operate in isolation, ‘networked self-awareness’ considers the effect of actions and interactions of other systems in the environment.

In the DCI group we study self-aware and networked self-aware processes and how they can support us to overcome the challenges of the future.

Emergent behaviour and self-organisation

Systems operating in a shared environment often interact with each other even when this is not intended. This can lead to behaviour and results that were not intended or even conceived at design time. Nevertheless, we have to make sure that systems do not develop behaviour that has a negative effect on themselves or on the environment.

In the DCI group, we study and develop approaches to steer self-organising and emergent behaviour in order to optimise the outcome for all systems in the environment.

Collective and swarm intelligence

Together we can do more. While we are interested in autonomous intelligent systems, we are also interested in their collaborative abilities. Bringing their knowledge together allows the individual to get a wider and deeper understanding. By distributing knowledge across a wide area and a large number of devices, we can tackle larger problems than individual systems could overcome.

In the DCI group we explore approaches enabling individual systems to incorporate information from other systems within their own knowledge-base.