Dr. Thorsten Weyer will present fundamental approaches for specifying highly automated self-learning systems at a meeting of the Requirements Engineering group of the Gesellschaft für Informatik (GI, German Informatics Society). The meeting will take place on November 29-30, 2018 in Nuremberg.
Whether autonomous road traffic, medicine, intelligent power grids or industry 4.0 – in all technology fields highly automated systems will be the key driver for digital transformation. Such systems often operate in highly dynamic, open contexts in order to meet overriding goals, e.g. in the case of coordinated overtaking processes of automated vehicle convoys ("platoon"). A central challenge in the development of highly automated systems is to equip the systems with the highest possible degree of resilience (in terms of reliability and robustness), i.e. to design and implement the systems in a way that they perform their tasks as reliably as possible during operation - even in the case of critical, unforeseen interference - without human intervention. Verified machine learning is a promising approach in this respect.
In his speech, Dr. Thorsten Weyer will look at selected challenges in the development of self-learning, highly automated systems and present basic approaches for the specification of such systems. These approaches form the foundation to systematically and comprehensibly realize highly automated systems, which have a high degree of resilience in long-term operation.
This year's topic of the group meeting is “RE and the Digital Transformation of Society”. Further information can be found on the website: http://fg-re.gi.de/treffen/treffen-2018.html (only in German).