ReBAR – Reducing Barriers for AI in (applied) Research

The ReBAR - Reducing Barriers for AI in (applied) Research - project directly addresses current challenges in the digital transformation of applied research. In the context of DLR's digitization strategy, this refers primarily to thrust 1 of the digitization initiative "Artificial Intelligence". In particular, the goal of "strengthening a broad user competence of AI methods in the institutes and facilities (IuE) where similar foundations and competencies are needed" is in the foreground.

Goals

The ReBAR project has two goals to enable the cross-program, engineering-centric application of artificial intelligence methods (especially in the context of machine learning): On the one hand, a prototypical framework is implemented that supports technical implementations of these methods in a user-centric way, and on the other hand, a user community is created at DLR. Currently, the use of these technologies requires a great understanding on the one hand of the domain being worked on (applications in L, R, V, E) as well as specific expertise in the AI methods being used. By creating a base-platform for all programs, barriers to understanding can be broken down in a concerted manner and methodological competence can be promoted. In addition, by structuring the methods, a scheme can also be developed to integrate other existing (conventional machine learning) and emerging methods (e.g., quantum-accelerated machine learning). This will also ensure connectivity for future technologies. By concretely addressing selected use cases from different DLR programs, the universal, cross-program applicability of the framework is ensured.

However, the framework radiates into further related directions (data, autonomous systems and cyber-physical engineering) and thus forms a basis for developments based on it.