Military Technology 06/2021

of this kind require considerable computing power. Therefore, supercom- puters capable of supporting new algorithms are needed. This is where the third major strand of the current innovation cycle comes in. Storage capacity is not intrinsically dependent on computing power, but it is es- sential to be able to exploit giant archives. This triad of technologies is the cornerstone of future air combat systems in terms of design, indus- trialisation, and operations. For example, today’s supercomputers, such as Leonardo’s DAVINCI-1, are artefacts that can be used to solve aircraft fluid dynamics problems, to calculate the effect of weight or thrust on a given component, to create the geometric profiles needed to reduce radar visibility, or to train the AI algorithms that will absorb and process even more information. Hierarchies and Architectures For an air combat system, in addition to the technological capability itself, all components must be optimally placed. The three key enablers can be located on the ground, onboard the fighter and its adjuncts, or on other air, sea, or land platforms. In principle, it would be possible to fully centralise storage and calculating capabilities in a single aggrega- tion infrastructure, which all objects in flight draw on and contribute to. The exponential growth in the amount of data, and the need to secure it, is pushing cloud architectures towards a more hierarchical structure on at least two levels. A proximity layer, in which raw data is collected and pre-selected – the ‘edge’ layer – so as not to congest aggregation infrastructures – the ‘central consolidation’ layer. This prevents the main infrastructure from being highly stressed during crises. Today, Microsoft, Google or Amazon networks in Europe support everyone’s data and Next-generation air combat systems promise unprecedented ca- pabilities and overwhelming superiority against any enemy. But what has been described so far at the design level depends on the successful development of three fundamental technologies: cloud computing, artificial intelligence (AI), and supercomputing. These complement and corroborate each other and represent the corner- stone of the entire architecture. Cloud Computing Cloud computing stems from the need to process data collected from multiple sources through a single storage space or device. It is a decen- tralised, delocalised storage capacity off the collection point. The absence of an on-board storage facility (or at least its reduction) and the possibility of storing data from multiple sensors without having to download it from the individual aircraft, are cloud’s main advantages. Of course, data then needs to be processed, classified, and possibly analysed, without being overwhelmed by the mass of incoming information. Therefore, cloud and AI are intrinsically linked. Extracting relevant information from large vol- umes of data is the base task assigned to AI applications. Calculating power is the key to exploiting the cloud. The exponential growth of data requires exponential calculating capabilities. Translating or turning data into useful information means making everything measurable. For ex- ample, a facial recognition programme can store large amounts of data that allow the software to ‘learn’ (machine learning) how to find the right information to identify a face. Once learned, the programme can reliably repeat the process very quickly, thus improving its accuracy. Algorithms Marco Giulio Barone is a political-military analyst based in Paris, and a regular contributor to Monch magazines. Marco Giulio Barone Collaborative Air Combat The Technology World Behind Aircraft Feature MT 6/2021 · 37 An artist’s impression of a network of aircraft around a 6 th -generation fighter, each in constant communication with groups and subgroups of other aircraft. These capabilities do not yet exist and are a vision of the future. (Photo: Indra)