XANDAR Impact in Aeronautical Informatics

Published by xandaradmin on

Recently, exiting changes are influencing the landscape of safety-critical avionic systems. Use-cases with lightweight vehicles emerge for the novel Urban Air Mobility (UAM) segment. Such vehicles demand functionality, which in part can only be realized efficiently with Artificial Intelligence (AI) / Machine Learning (ML). Additionally, paradigm shifts towards agile and DevOps practice are striven for in order to cope with challenging time-to-market requirements. However, the aviation domain is tightly regulated, with developments conducted using stringent processes. These are guided through rigorous quality and safety management based on traceable multi-level requirements, extensive documentation and huge efforts in verification.

Urban Air Mobility Concepts – DLR (CC BY-NC-ND 3.0)

Correctness-by-Design and foremost model-based software practices have lastingly shaped the engineering culture in the aviation domain. Nonetheless, especially with the recent changes, there still remains huge potential for further improvement.

Now in the XANDAR toolchain, the aforementioned is combined to ultimately advance the state of the art. Foremost, XANDAR aims to achieve X-by-Construction (XbC). In XbC, X substitutes multiple aviation relevant non-functional properties such as safety, security, and timing. By integrating the verification of both functional and non-functional requirements in the development process, XANDAR lowers the burden towards achieving XbC even further. But, the highly integrated and automated nature of the toolchain strives for more. As verification and validation mechanism are directly baked into the design process, XANDAR assists in complying with recent challenges of safety-critical system developments. One of them being increasingly agile practices, which is addressed with the ability to efficiently support changing requirements. This is considered even for non-functional properties. At the same time, it is investigated how safety can be improved for the use of novel AI/ML software components. Exemplary, monitoring of the Operation Design Domain (ODD) will be one considered aspect in the toolchain.

In summary, recent and exciting opportunities in the aviation domain also introduce challenges towards the development of safety-critical systems. Focal point for the XANDAR toolchain is geared towards addressing many of these challenges. This highlights the potential for the XANDAR approach and will drive the further advancement of the project.

Categories: News