Blog

In terms of size, complexity and industrial capacity required, the most advanced test launched in Italy involving the combined use of supercomputing and artificial intelligence is underway in the defence sector. It is called ‘Michelangelo Dome’ and is being conducted by Leonardo.

By exploiting big data, predictive algorithms, large proprietary language models, neural networks, generative artificial intelligence, high-performance computing, remote command and control systems and intelligent sensors located on ships, aircraft, satellites and ground infrastructure, Italy’s leading defence and aerospace company has designed a ‘security dome’ capable of detecting, tracking and neutralising old and new threats. From attacks conducted by aircraft or missiles – including hypersonic missiles and swarms of drones – to those launched by sea or land, to hybrid attacks launched from the digital world. ‘Michelangelo Dome’ is programmed to anticipate enemy moves, optimise response times and automatically identify the most effective countermeasures. And thus enhance European and NATO countries’ defence capabilities.

‘This project,’ explains Simone Ungaro, Co-general manager strategy & innovation at Leonardo, ‘constitutes the integrated air defence system we are developing as part of multi-domain solutions. It is a challenge that requires infinite computational and artificial intelligence capabilities to support command and control platforms. The system,’ Ungaro continues, ‘has to exchange huge amounts of data in real time, analyse them and turn them into immediate operational decisions. That is why we are working to bring computing power not only into our data centres, but also into sensors, on board ships and aircraft, on satellites, and on every active defence system in the field’.

Leonardo’s transformation

There are two key technologies that Leonardo has at its disposal to tackle this new, complicated industrial challenge. Artificial intelligence and supercomputing. The company has been able to take up the challenge thanks to as many strategic infrastructures at its disposal: the Davinci-1 supercomputer – capable of performing 5 million billion operations per second – and the Leonardo innovation Labs, technological hubs created to do research and development on frontier technologies. Two levers present in Leonardo thanks to a visionary process of corporate transformation launched in 2019 by the current CEO Roberto Cingolani, which has changed the face of the company forever. Today, 200 employees are dedicated to the development of the supercomputing infrastructure. While more than 2,000 are already accessing the Davinci-1 for engineering simulations, AI-based predictive analysis, satellite image studies and the development of new technologies.

Installed in Genoa, the Davinci-1 supercomputer will soon be upgraded. The company will soon inaugurate the Davinci-2, which will have 30 petabytes of memory and 20 Petaflops of computing power. Power increases to 1.2 Exaflops for 8-bit operations, those typical of AI applications. ‘Around this ecosystem that combines computing power, cloud resources and artificial intelligence,’ Ungaro explains, ‘Leonardo has developed distinctive skills. It is no coincidence that we are part of the shortlist of companies that have finalised the Italian proposal for the assignment of one of the five AI Gigafactories that the EU wants to build’.