Immo Colonius

Background

About Immo Colonius

Since coming into contact with modelling of game adversaries, i was captivated by the broad idea of artificial intelligence. Especially the approach to model knowledge in a form that is usable for automatic inference systems (facilitating decision mechanisms) and on the same side still understandable for humans appealed to me. Formal representations in logic predicates were a good start on which I build for spatial / temporal reasoning in my current research. This is the driving challenge for my scientific career so far. Impressed by the usability of abstraction to reduce clutter in information I choose that to be the topic of by diploma theses. I've shown that it is possible to reason efficiently by logic deduction, even when presented with noisy and incomplete data sets for a location task of an agent in a virtual environment. Besides the work on reasoning methods and the visualisation of logistic processes I was keen on broadening my view on different aspects of Artificial Intelligence and international cooperation and accepted a position in the SFB TR8 - Spatial Cognition. I acquired knowledge about different sensory prerequisites for transfer of information and had to communicate challenges and findings to experts from a variety of professions, for example doctors, communications technicians and EU commissions, making me well versed in switching granularities for explanations. Besides a multiple month evaluation period in Rome, Italy, frequent contact with a large international team schooled me in large team collaborations. Currently I formalised the "Convention on the International Regulations for Preventing Collisions at Sea" into a symbolic representation and deployed pattern matching methods based on logic reasoning with a very efficient implementation, enabling near real time situation analysis and provability of results as an inherent feature. This has been shown functional within international conferences and in my upcoming dissertation.