DMI Think Piece: Relevance of ethical design principles for future Mission Systems

In times of exponential data production in an increasingly interconnected, globalized and digitalized world, we believe it is crucial to leverage the responsible use data. Our mission is to support high-tech industries such as aerospace and defence, industrial production, transport and logistics to make full use of possibilities of real-time analysis of large amounts of data.


Data Machine Intelligence Solutions has been selected by Airbus Defence and Space for the German Non-Traditional Players innovation initiative to be part of the Future Combat Air Systems Program. FCAS is Europe’s largest and most ambitious military aeronautics development programme, launched in 2017 by Germany, France and Spain.


In this program a multi-dimensional Combat Cloud, integrating data in real time from various combat systems, will prepare, process, and present user-friendly and relevant information and up-to-date analyses, ensuring optimal decisions for the mission commander. This Combat Cloud is meant to operate on various levels: in a context where combat aircrafts will team with remote carrier in close cooperation with allied forces, they will be enabled to re-evaluate the mission in real time thanks to the constant data stream and analysis of land, maritime, space, air and cyber forces. The use of AI for combat missions is diverse and useful: collecting, stocking, sharing and analysing of input data from various platforms, reconnaissance, detection and response to cyberthreats, situational awareness and physical threat detection, mission planification, live re-planning of operations and coordination of drone swarms are some examples of the possibilities of the use of AI in combat missions.


The aim is that this multi-dimensional Combat Cloud significantly improves the “observe, orient, decide and act” loop to ensure more successful, precise and efficient

missions thanks to fast access to trusted data.


Most importantly, this Combat Cloud aims at helping human-driven decisions and is meant to be kept under the full control of a human operator at any time. Therefore, a cross platform mission system must allow large data analyses with a speed and efficiency that could not be matched by a human being. As a result, the human operator can make the most relevant and informed decision in the highly evolving and intense contexts of combat missions. In more technical terms, such a cloud-based mission system with a layered interface must allow the user to see the processing, analysis and decision-making of the AI live at any given point in time. Thus the interface will need to become effectively a digital assistant to the pilot.


Conscious of the ethical aspect of creating AI assistance that will process information, analyse and deliver propositions for combat missions, it needs to be guaranteed that the is able to fully understand the systems advice to him. The AI analyses might then be referred to as “suggestions” that will always have to be approved by the pilot.


The revolutionary aspect of “ethical design” will be the following: the AI interface will be trained to obey a list of behaviours compliant with e.g Humanitarian Law and deducted Rules of engagement helping to comply with those norms without risking mission success. In other words, the AI will be trained to help the pilot avoiding collateral damage. Moreover, pilots will have the possibility to check every step of the way how the interface arrived at a certain conclusion at any point in time. In a second step the mission systems may perform after-mission data analyses, to learn lessons of previous missions and re-evaluate strategies in the long run if needed.


Data Machine Intelligence Solutions is working since 2021 with the AI experts of Fraunhofer FKIE for Airbus Defence and Space to develop an Explainable AI (XAI) demonstrator. The project is realized jointly with the FCAS Expert Commission on the responsible use of new technologies (https://www.fcas-forum.eu/).






61 Ansichten

Aktuelle Beiträge

Alle ansehen